15 days14136.6\nNeonatal gender\n Male18848.8 Female19751.2\nBirth weight\n High birth weight8321.5 Normal birth weight16242.1 Low birth weight14036.4\nGestational age\n Premature6015.6 Full term32584.4\nNeonatal hospital stay\n ≤ 15 days23561 > 15 days15039\nAdmission diagnosis\n Infection6516.9 Respiratory problems11529.9 Prematurity5915.3 Gastrointestinal problems389.9 Jaundice4712.2 Neurological problems215.5 Cardiological problems266.8 Others143.5\n\nSocio-demographic characteristics of study participants (n = 385)\n[SUBTITLE] Explanatory factor analysis for subscales of parental satisfaction with NICU-care service [SUBSECTION] Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool\nBefore assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool\n[SUBTITLE] Reliability of items of EMPATHIC-N tool [SUBSECTION] The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable\nThe total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable\n[SUBTITLE] Parental satisfaction level with NICU-care service [SUBSECTION] Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services\nOverall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services\n[SUBTITLE] Factors associated with the overall parental satisfaction of NICU care service [SUBSECTION] In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model\nIn this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"There was a low level of parental satisfaction with neonatal intensive care unit service. Among the dimensions of EMPATHIC-N, the lowest parental satisfaction score was in the care and treatment while the highest parental satisfaction score was in the organization dimension. As a result, health professionals and hospital administrators should collaborate to improve NICU services to provide high-quality service and satisfy parents."},"other_sections_titles":{"kind":"list like","value":["Methods and materials","Study area and period","Inclusion and exclusion criteria","Sample size and sampling technique","Data collection instrument and procedures","Data quality assurance","Ethical consideration","Data entry and analysis","Operational definitions","Explanatory factor analysis for subscales of parental satisfaction with NICU-care service","Reliability of items of EMPATHIC-N tool","Parental satisfaction level with NICU-care service","Factors associated with the overall parental satisfaction of NICU care service","Limitations of the study"],"string":"[\n \"Methods and materials\",\n \"Study area and period\",\n \"Inclusion and exclusion criteria\",\n \"Sample size and sampling technique\",\n \"Data collection instrument and procedures\",\n \"Data quality assurance\",\n \"Ethical consideration\",\n \"Data entry and analysis\",\n \"Operational definitions\",\n \"Explanatory factor analysis for subscales of parental satisfaction with NICU-care service\",\n \"Reliability of items of EMPATHIC-N tool\",\n \"Parental satisfaction level with NICU-care service\",\n \"Factors associated with the overall parental satisfaction of NICU care service\",\n \"Limitations of the study\"\n]"},"other_sections_texts":{"kind":"list like","value":["[SUBTITLE] Study area and period [SUBSECTION] A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\nA cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\n[SUBTITLE] Inclusion and exclusion criteria [SUBSECTION] Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\nParents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\n[SUBTITLE] Sample size and sampling technique [SUBSECTION] The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\nThe sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\n[SUBTITLE] Data collection instrument and procedures [SUBSECTION] The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\nThe Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\n[SUBTITLE] Data quality assurance [SUBSECTION] To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\nTo ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\n[SUBTITLE] Ethical consideration [SUBSECTION] Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\nDebre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\n[SUBTITLE] Data entry and analysis [SUBSECTION] Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\nData were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\n[SUBTITLE] Operational definitions [SUBSECTION] Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\nSatisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].","A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.","Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.","The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.","The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).","To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.","Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.","Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.","Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].","Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool","The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable","Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services","In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model","The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation."],"string":"[\n \"[SUBTITLE] Study area and period [SUBSECTION] A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\\nA cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\\n[SUBTITLE] Inclusion and exclusion criteria [SUBSECTION] Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\\nParents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\\n[SUBTITLE] Sample size and sampling technique [SUBSECTION] The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\\nThe sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\\n[SUBTITLE] Data collection instrument and procedures [SUBSECTION] The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\\nThe Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\\n[SUBTITLE] Data quality assurance [SUBSECTION] To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\\nTo ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\\n[SUBTITLE] Ethical consideration [SUBSECTION] Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\\nDebre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\\n[SUBTITLE] Data entry and analysis [SUBSECTION] Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\\nData were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\\n[SUBTITLE] Operational definitions [SUBSECTION] Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\\nSatisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\",\n \"A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\",\n \"Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\",\n \"The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\",\n \"The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\",\n \"To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\",\n \"Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\",\n \"Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\",\n \"Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\",\n \"Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\\n\\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\\n1\\n\\n2\\n\\n3\\n\\n4\\n\\nInformation\\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\\nCare & Treatment\\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\\nParental Participation\\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\\nOrganization\\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\\nProfessional Attitude\\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\\n\\nFactor loadings and identified components’ of EMPATHIC-N tool\",\n \"The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\\n\\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\\n\\nReliability of items of parental satisfaction with NICU-care service\\n*Significant value (p < 0.001), c = not computable\",\n \"Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\\n\\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\\n\\nParents’ overall and dimensional satisfaction with NICU-care services\",\n \"In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\\n\\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\\nGender\\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\\nResidency\\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\\nParental hospital stay\\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\\nBirth weight\\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\\nGestational age\\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\\n\\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\\n*= p-value < 0.05 1 = reference BWt-birth weight\\nNote: p-values were extracted from the multivariate logistic regression model\",\n \"The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods and materials","Study area and period","Inclusion and exclusion criteria","Sample size and sampling technique","Data collection instrument and procedures","Data quality assurance","Ethical consideration","Data entry and analysis","Operational definitions","Results","Explanatory factor analysis for subscales of parental satisfaction with NICU-care service","Reliability of items of EMPATHIC-N tool","Parental satisfaction level with NICU-care service","Factors associated with the overall parental satisfaction of NICU care service","Discussion","Limitations of the study","Conclusion"],"string":"[\n \"Introduction\",\n \"Methods and materials\",\n \"Study area and period\",\n \"Inclusion and exclusion criteria\",\n \"Sample size and sampling technique\",\n \"Data collection instrument and procedures\",\n \"Data quality assurance\",\n \"Ethical consideration\",\n \"Data entry and analysis\",\n \"Operational definitions\",\n \"Results\",\n \"Explanatory factor analysis for subscales of parental satisfaction with NICU-care service\",\n \"Reliability of items of EMPATHIC-N tool\",\n \"Parental satisfaction level with NICU-care service\",\n \"Factors associated with the overall parental satisfaction of NICU care service\",\n \"Discussion\",\n \"Limitations of the study\",\n \"Conclusion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Neonatal intensive care units (NICUs) are areas that require careful risk management with a wide range of neonatal care services [1–3]. It necessitates high-cost and efficient critical care delivery with a multidisciplinary team approach that focuses on preventive strategies for improved outcomes [4–6]. Parental tension and emotions are high when their child is admitted to a neonatal intensive care unit (NICU) due to serious illnesses [7, 8].\nSatisfaction is a belief and attitude about a specific service provision of an institution. Parental and patient satisfaction has become a well-established outcome indicator and tool for assessing a healthcare system’s quality, as well as input for developing strategies and providing accessible, sustainable, economical, as well as acceptable patient care [7, 9, 10]. Parental satisfaction reflects the balance between their expectations of ideal care and their perception of real and available care [3, 11, 12]. It is also one of the objectives and missions of every health care center that gives NICU care service [10, 11].\nParent and patient satisfaction has become an important aspect of hospital management initiatives for quality assurance and accreditation. Also, parental involvement has an important role in the delivery of high-quality care, ranging from assisting with activities of daily living to being directly involved in important health care decisions [13, 14]. However, parental participation may not always be possible, but, effective communication will reduce the effects of crises and improve parental satisfaction [11, 15].\nEthiopia has been working to enhance its healthcare delivery systems by focusing on quality healthcare service giving special attention to mothers and children. The Ethiopian Health Service Alliance for Quality has agreed that the initial priority area would be self-motivated and transparent partnerships that stimulate innovation in health care quality management and learning across all hospitals [16].\nGiven the scarcity of studies on parental satisfaction with NICU care in Ethiopia, as well as clinical observations of parents complaining about NICU care, it is critical to determine the level of parental satisfaction. So, this study aimed to identify the level of parental satisfaction and associated factors with NICU care service in Debre Tabor Comprehensive Specialized Hospital (DTCSH). Also, this study helps the administrators to understand the deficiencies of the hospital’s NICU services, and then propose corresponding improvement strategies.","[SUBTITLE] Study area and period [SUBSECTION] A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\nA cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\n[SUBTITLE] Inclusion and exclusion criteria [SUBSECTION] Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\nParents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\n[SUBTITLE] Sample size and sampling technique [SUBSECTION] The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\nThe sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\n[SUBTITLE] Data collection instrument and procedures [SUBSECTION] The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\nThe Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\n[SUBTITLE] Data quality assurance [SUBSECTION] To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\nTo ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\n[SUBTITLE] Ethical consideration [SUBSECTION] Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\nDebre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\n[SUBTITLE] Data entry and analysis [SUBSECTION] Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\nData were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\n[SUBTITLE] Operational definitions [SUBSECTION] Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\nSatisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].","A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.","Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.","The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n={\\left({Z}_{\\frac{a}{2}}\\right)}^{2}P{(1-P)/d}^{2}$$\\end{document}\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\nN = 385.\nBy considering a 5% non-respondent rate the final sample size was 405.\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.","The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).","To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.","Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.","Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.","Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\nLow birth weight: neonates with a birth weight of lower than 250 g [25].","This study was conducted on a total of 385 parents, with a 95.06% response rate. The majority of the parents were mothers 224 (58.2%), whereas 322 (83.6%) were married. Full-term newborns 325(84.4%) and those with respiratory issues 115 (29.9%) had the highest number of NICU admissions (Table 1).\n\nTable 1Socio-demographic characteristics of study participants (n = 385)VariablesFrequency (n)Percentage (%)\nParental gender\n Mother22458.2 Father16141.8\nAge\n 18–2413234.3 25–3916643.1 40 and above8722.6\nMarital status\n Married32283.6 Not married6316.4\nResidency\n Urban18748.6 Rural19851.4\nEducational level\n No formal education11930.9 Elementary school9223.9 Secondary school8321.6 Collage and above9123.6\nProfession\n Housewife7018.2 Farmer10627.5 Student5915.3 Gov’t employ7218.7 Private employ7820.3\nParental hospital stay\n ≤ 15 days24463.4 > 15 days14136.6\nNeonatal gender\n Male18848.8 Female19751.2\nBirth weight\n High birth weight8321.5 Normal birth weight16242.1 Low birth weight14036.4\nGestational age\n Premature6015.6 Full term32584.4\nNeonatal hospital stay\n ≤ 15 days23561 > 15 days15039\nAdmission diagnosis\n Infection6516.9 Respiratory problems11529.9 Prematurity5915.3 Gastrointestinal problems389.9 Jaundice4712.2 Neurological problems215.5 Cardiological problems266.8 Others143.5\n\nSocio-demographic characteristics of study participants (n = 385)\n[SUBTITLE] Explanatory factor analysis for subscales of parental satisfaction with NICU-care service [SUBSECTION] Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool\nBefore assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool\n[SUBTITLE] Reliability of items of EMPATHIC-N tool [SUBSECTION] The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable\nThe total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable\n[SUBTITLE] Parental satisfaction level with NICU-care service [SUBSECTION] Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services\nOverall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services\n[SUBTITLE] Factors associated with the overall parental satisfaction of NICU care service [SUBSECTION] In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model\nIn this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model","Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\n\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\n1\n\n2\n\n3\n\n4\n\nInformation\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\nCare & Treatment\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\nParental Participation\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\nOrganization\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\nProfessional Attitude\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\n\nFactor loadings and identified components’ of EMPATHIC-N tool","The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\n\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\n\nReliability of items of parental satisfaction with NICU-care service\n*Significant value (p < 0.001), c = not computable","Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\n\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\n\nParents’ overall and dimensional satisfaction with NICU-care services","In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\n\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\nGender\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\nResidency\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\nParental hospital stay\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\nBirth weight\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\nGestational age\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\n\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\n*= p-value < 0.05 1 = reference BWt-birth weight\nNote: p-values were extracted from the multivariate logistic regression model","In this study, the total average parental satisfaction score with NICU service was 47.8%. This finding is nearly similar to a study done in Ethiopia (50%) [26]. Contrary to this finding, studies done in Norway (76%) [7] and Greece (99%) [27] had a higher parental satisfaction score with NICU service.\nThe study found that parental satisfaction with NICU service was 50.4% in the information subscale, 36.9% in the care and treatment subscale, 50.1% in the parental participation subscale, 59.0% in the organization subscale, and 48.6% in the professional attitude subscale. Parental satisfaction was lowest in the care and treatment subscale. A study in Italy [20] and South Africa [28] on the other hand, discovered the lowest parental satisfaction score in the professional attitude and parental participation subscales respectively. This disparity could be attributed to the NICU’s lack of professional and medical resources to treat and care for neonates in this situation.\nAccording to this study, mothers, parents from rural regions, parents of neonates who stayed less than 15 days in the hospital, parents of neonates with normal birth weight, and parents of full-term neonates were all more satisfied with NICU services than their counterparts.\nThis study showed that mothers were more satisfied with NICU service than fathers. Similar studies conducted in Italy [20], Israel [11], and Greece [29] also showed that mothers were more satisfied than fathers in NICU service. The possible explanation might be women are allowed to spend more time in the NICU and participate in the care of their newborns and cultivate more relationships with medical caregivers than fathers.\nAlso in this study parents who came from rural areas were more satisfied than parents from urban areas. This founding was also similar to a study done in Greece [30] found that parents from rural areas were more satisfied. The probable reason for this might be parents from rural areas may have a low-level awareness of the hospital, expectations, and demand for NICU service as compared with actual practice.\nParents of normal birth weight and full-term neonates were more satisfied with NICU service than parents of low birth weight and preterm neonates in this study. In addition, a Norwegian study [7] found that parents of newborns with normal birth weights were more satisfied than those with low birth weights. Reasonably, parents of full-term and normal-weight neonates are likely to face unexpected concerns, which may lead to greater satisfaction with positive outcomes.\nIn this study, parents of neonates who stayed less than 15 days in the hospital were more satisfied with NICU service than parents who stayed 15 or more days. Similarly, studies done in Ethiopia [26] and Iran [10] found parents who stayed for a short period in the NICU were more satisfied. The possible reason for this might be that parents with short-stay are less likely to see their neonatal serious conditions that make emotional and care mismanagements.\n[SUBTITLE] Limitations of the study [SUBSECTION] The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.\nThe study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.","The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.","There was a low level of parental satisfaction with neonatal intensive care unit service. Among the dimensions of EMPATHIC-N, the lowest parental satisfaction score was in the care and treatment while the highest parental satisfaction score was in the organization dimension. As a result, health professionals and hospital administrators should collaborate to improve NICU services to provide high-quality service and satisfy parents."],"string":"[\n \"Neonatal intensive care units (NICUs) are areas that require careful risk management with a wide range of neonatal care services [1–3]. It necessitates high-cost and efficient critical care delivery with a multidisciplinary team approach that focuses on preventive strategies for improved outcomes [4–6]. Parental tension and emotions are high when their child is admitted to a neonatal intensive care unit (NICU) due to serious illnesses [7, 8].\\nSatisfaction is a belief and attitude about a specific service provision of an institution. Parental and patient satisfaction has become a well-established outcome indicator and tool for assessing a healthcare system’s quality, as well as input for developing strategies and providing accessible, sustainable, economical, as well as acceptable patient care [7, 9, 10]. Parental satisfaction reflects the balance between their expectations of ideal care and their perception of real and available care [3, 11, 12]. It is also one of the objectives and missions of every health care center that gives NICU care service [10, 11].\\nParent and patient satisfaction has become an important aspect of hospital management initiatives for quality assurance and accreditation. Also, parental involvement has an important role in the delivery of high-quality care, ranging from assisting with activities of daily living to being directly involved in important health care decisions [13, 14]. However, parental participation may not always be possible, but, effective communication will reduce the effects of crises and improve parental satisfaction [11, 15].\\nEthiopia has been working to enhance its healthcare delivery systems by focusing on quality healthcare service giving special attention to mothers and children. The Ethiopian Health Service Alliance for Quality has agreed that the initial priority area would be self-motivated and transparent partnerships that stimulate innovation in health care quality management and learning across all hospitals [16].\\nGiven the scarcity of studies on parental satisfaction with NICU care in Ethiopia, as well as clinical observations of parents complaining about NICU care, it is critical to determine the level of parental satisfaction. So, this study aimed to identify the level of parental satisfaction and associated factors with NICU care service in Debre Tabor Comprehensive Specialized Hospital (DTCSH). Also, this study helps the administrators to understand the deficiencies of the hospital’s NICU services, and then propose corresponding improvement strategies.\",\n \"[SUBTITLE] Study area and period [SUBSECTION] A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\\nA cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\\n[SUBTITLE] Inclusion and exclusion criteria [SUBSECTION] Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\\nParents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\\n[SUBTITLE] Sample size and sampling technique [SUBSECTION] The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\\nThe sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\\n[SUBTITLE] Data collection instrument and procedures [SUBSECTION] The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\\nThe Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\\n[SUBTITLE] Data quality assurance [SUBSECTION] To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\\nTo ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\\n[SUBTITLE] Ethical consideration [SUBSECTION] Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\\nDebre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\\n[SUBTITLE] Data entry and analysis [SUBSECTION] Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\\nData were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\\n[SUBTITLE] Operational definitions [SUBSECTION] Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\\nSatisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\",\n \"A cross-sectional study was conducted in DTCSH which is a public hospital established in 1934 and located in the South Gondar Zone of Amara Regional State of Ethiopia. It is 97 km to the southwest of Bahir Dar, the capital city of Amara Regional State. According to the 2007 census, the total population of Debre tabor town was 155,596. It has a latitude and longitude of 11051N3801’E11.8500 N 38.0170E with an elevation of 2,706 m (8878ft) above sea level. The hospital provides neonatal intensive care unit service with five separate NICU rooms. According to the hospital’s 2017 report, 1159 neonates were admitted to the NICU, but according to evidence from a chart review in 2020, 1489 neonates were admitted [17]. This study was conducted on parents of neonates who were admitted to NICU at DTCSH from November 05, 2021, to April 30, 2022.\",\n \"Parents whose neonate was discharged from NICU or transferred to high dependency neonatal ward, parents who can read and write or understand the Amharic language, and neonatal stay in NICU of less than three months were included in the study. Whereas, parents with neonatal death in NICU and parents with neonatal admissions of less than 24 h were excluded from the study.\",\n \"The sample size was determined by using single proportion population formula taking 50% (P) of the proportion using a 95% confidence interval and 5% margin of error (d). The sample size was determined using the following formula.\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n={\\\\left({Z}_{\\\\frac{a}{2}}\\\\right)}^{2}P{(1-P)/d}^{2}$$\\\\end{document}\\nn= [(1.96)2 0.5(1-0.5)2] / (0.05)2.\\nN = 385.\\nBy considering a 5% non-respondent rate the final sample size was 405.\\nThe data were collected from all consecutive parents who met the inclusion criteria until the intended sample size was achieved.\",\n \"The Amharic version anonymous questionnaire was used for data collection after translating the English version of the EMPATHIC-N tool by three language experts and then back to English by the other three experts to ensure that the translation was correct. The tool’s content validity was also examined and guaranteed by members of the Anesthesia department’s research committee.\\nThe tool has been widely used to assess parental satisfaction with NICU care services, and it has strong reliability and validity, with reliability (Cronbach’s) values of the domains ranging from 0.82 to 0.95 [18–20]. There are five domains in the EMPATHIC-N tool: Information (12 questions with a six-point Likert scale); Care & Treatment (17 questions with a ten-point Likert scale); Parental Participation (8 questions with a six-point Likert scale); Organization (8 questions with a six-point Likert scale); and Professional Attitude (12 questions with a six-point Likert scale) [18, 21, 22].\\nThe data were collected by three anesthetists on the day of neonatal discharge using an adopted Dutch instrument of Empowerment of Parents in The Intensive Care-Neonatology (EMPATHIC-N).\",\n \"To ensure the quality of data, pre-testing of the data collection tool (the questionnaire) was done on 5% of study parents from Felege Hiwot Comprehensive Specialized Hospital who were not included in the main study. The training was given to data collectors; data were collected and properly filled in the prepared format. Supervision was made throughout the data collection period to make sure the accuracy, clarity, and consistency of the collected data.\",\n \"Debre Tabor University provided ethical clearance, and each parent was given written informed consent after being briefed about the purpose study.\",\n \"Data were cleaned, coded, and entered into Epidata version 4.2 and exported to SPSS version 23 for statistical analysis. The adopted EMPATHIC-N instrument was validated using analysis (validity, reliability, standard factor loadings, and factor analysis). Cronbach’s alpha was used to determine the reliability and validity of the tool. Explanatory factor analysis was done to test how well the measured variables represent the number of constructs and to identify relationships between the measured items. The inter-item correlation was used to assess the relationship between items on the same scale, whereas item-discriminant validity was used to assess the relationship between scales. After categorizing the overall mean parental satisfaction score, independent variables were analyzed using binary logistic regression with parental satisfaction with NICU care service. Variables from the bivariable logistic regression with a p-value of 0.2 were fitted to a multivariable logistic regression, and certain variables were included in the model with their clinical importance. Both crude odds ratio (COR) in bivariable logistic regression and adjusted odds ratio (AOR) in multivariable logistic regression with the corresponding 95% Confidence interval were calculated to show the strength of association. In multivariable logistic regression analysis, variables with a p-value of < 0.05 were considered statistically significant. The Mann–Whitney test was used to determine the influencing factors of the parental satisfaction domains, while the Hosmer–Lemeshow goodness of fit test was performed to ensure that the analysis model was appropriate.\",\n \"Satisfied: parents who scored greater than or equal to the overall mean EMPATHIC-N values were considered satisfied.\\nDissatisfied: parents who scored less than the overall mean EMPATHIC-N values were considered dissatisfied.\\nHigh birth weight: neonates with a birth weight of more than 400 g [23].\\nNormal birth weight: neonates with a birth weight range from 250 to 400 g [24].\\nLow birth weight: neonates with a birth weight of lower than 250 g [25].\",\n \"This study was conducted on a total of 385 parents, with a 95.06% response rate. The majority of the parents were mothers 224 (58.2%), whereas 322 (83.6%) were married. Full-term newborns 325(84.4%) and those with respiratory issues 115 (29.9%) had the highest number of NICU admissions (Table 1).\\n\\nTable 1Socio-demographic characteristics of study participants (n = 385)VariablesFrequency (n)Percentage (%)\\nParental gender\\n Mother22458.2 Father16141.8\\nAge\\n 18–2413234.3 25–3916643.1 40 and above8722.6\\nMarital status\\n Married32283.6 Not married6316.4\\nResidency\\n Urban18748.6 Rural19851.4\\nEducational level\\n No formal education11930.9 Elementary school9223.9 Secondary school8321.6 Collage and above9123.6\\nProfession\\n Housewife7018.2 Farmer10627.5 Student5915.3 Gov’t employ7218.7 Private employ7820.3\\nParental hospital stay\\n ≤ 15 days24463.4 > 15 days14136.6\\nNeonatal gender\\n Male18848.8 Female19751.2\\nBirth weight\\n High birth weight8321.5 Normal birth weight16242.1 Low birth weight14036.4\\nGestational age\\n Premature6015.6 Full term32584.4\\nNeonatal hospital stay\\n ≤ 15 days23561 > 15 days15039\\nAdmission diagnosis\\n Infection6516.9 Respiratory problems11529.9 Prematurity5915.3 Gastrointestinal problems389.9 Jaundice4712.2 Neurological problems215.5 Cardiological problems266.8 Others143.5\\n\\nSocio-demographic characteristics of study participants (n = 385)\\n[SUBTITLE] Explanatory factor analysis for subscales of parental satisfaction with NICU-care service [SUBSECTION] Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\\n\\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\\n1\\n\\n2\\n\\n3\\n\\n4\\n\\nInformation\\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\\nCare & Treatment\\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\\nParental Participation\\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\\nOrganization\\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\\nProfessional Attitude\\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\\n\\nFactor loadings and identified components’ of EMPATHIC-N tool\\nBefore assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\\n\\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\\n1\\n\\n2\\n\\n3\\n\\n4\\n\\nInformation\\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\\nCare & Treatment\\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\\nParental Participation\\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\\nOrganization\\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\\nProfessional Attitude\\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\\n\\nFactor loadings and identified components’ of EMPATHIC-N tool\\n[SUBTITLE] Reliability of items of EMPATHIC-N tool [SUBSECTION] The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\\n\\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\\n\\nReliability of items of parental satisfaction with NICU-care service\\n*Significant value (p < 0.001), c = not computable\\nThe total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\\n\\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\\n\\nReliability of items of parental satisfaction with NICU-care service\\n*Significant value (p < 0.001), c = not computable\\n[SUBTITLE] Parental satisfaction level with NICU-care service [SUBSECTION] Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\\n\\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\\n\\nParents’ overall and dimensional satisfaction with NICU-care services\\nOverall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\\n\\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\\n\\nParents’ overall and dimensional satisfaction with NICU-care services\\n[SUBTITLE] Factors associated with the overall parental satisfaction of NICU care service [SUBSECTION] In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\\n\\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\\nGender\\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\\nResidency\\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\\nParental hospital stay\\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\\nBirth weight\\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\\nGestational age\\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\\n\\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\\n*= p-value < 0.05 1 = reference BWt-birth weight\\nNote: p-values were extracted from the multivariate logistic regression model\\nIn this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\\n\\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\\nGender\\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\\nResidency\\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\\nParental hospital stay\\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\\nBirth weight\\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\\nGestational age\\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\\n\\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\\n*= p-value < 0.05 1 = reference BWt-birth weight\\nNote: p-values were extracted from the multivariate logistic regression model\",\n \"Before assessing parental satisfaction levels, confirmatory factor analysis was used to ensure that the measured variables accurately represented the constructs. The factor correlation matrix was used in the analysis. KMO and Bartlett’s tests of sphericity were used to check the correlation of measurement variables, with a KMO value of 0.875 and Bartlett’s tests of sphericity (p = 0.00) respectively. The correlation matrix was checked for commonality extraction, and all item values were larger than 0.3. The parallel analysis determined the number of item components (factors), and two new components for parental participation, three new components for information and organization, and four new components for Care & Treatment and professional attitude of eigenvalues met the criteria, and the Varimax component correlation matrix was an appropriate model (Table 2).\\n\\nTable 2Factor loadings and identified components’ of EMPATHIC-N toolItemFactors\\n1\\n\\n2\\n\\n3\\n\\n4\\n\\nInformation\\n What level of doctor’s information do you have regarding the child’s expected health outcomes?0.752 How satisfied are you with the physicians’ and nurses’ information similarity?0.704 How are you satisfied with doctors’ and nurses’ honesty in providing information?0.59 How satisfied are you with daily discussions with doctors and nurses about your child’s care and treatment?0.58 How understandable was the information provided by the doctors and nurses?0.763 How satisfied were you with the correct information when the child’s physical condition deteriorated?0.4280.657 How satisfied were you with the clear answers to your questions?0.653 How clear is the doctor’s information about the consequences of the child’s treatment?0.533 To what extent do you receive clear information about the examinations and tests?0.713 To what extent the information brochure received was complete and clear?0.641 How much clear information is given regarding a child’s illness?0.626 Level of received understandable information about the effects of the drugs?0.606\\nCare & Treatment\\n Level of child’s comfort taken into account by the doctors and nurses?0.826 The extent of satisfaction During acute situations on availability of nurses to support?0.822 Level of team alertness to the prevention and treatment of pain of neonate?0.797 Level of care taken to nurses while in the incubator/bed?0.6340.501 Level of correct medication always administered on time?0.6310.402 level of emotional support that has been provided?0.84 The doctors and nurses responded well to our own needs0.809 Transferals of care from the neonatal intensive care unit staff to colleagues in the high-care unit or pediatric ward had gone well?0.782 Every day we knew who of the doctors and nurses was responsible for our child.0.695 How closely did doctors and nurses collaborate during work?0.5780.466 Level of a common goal: to provide the finest care and treatment for our child and ourselves.0.774 Level of physicians and nurses paid close attention to our child’s development?0.756 The team as a whole was concerned for our child and you.0.698 Our child’s requirements were met promptly0.455 The extent of doctors’ and nurses’ professional knowledge of what they are doing?0.847 How satisfied are you with the doctors’ and nurses’ understanding of the child’s medical history at the time of admission?0.765 How satisfied are you with the rapid actions taken by doctors and nurses when a child’s condition deteriorated?0.723\\nParental Participation\\n How involved are you in making decisions about our child’s care and treatment?0.803 The nurses had trained us on the specific aspects of newborn care.0.796 We were encouraged to stay close to our children.0.714 Before discharge, the care for our child was once more discussed with us.0.617 Even during intensive procedures, we could always stay close to our child.0.841 The nurses stimulated us to help in the care of our child0.837 The nurses helped us in the bonding with our child0.5320.575 We had confidence in the team0.561\\nOrganization\\n The Neonatology unit made us feel safe0.885 There was a warm atmosphere in the Neonatology unit without hostility0.805 The Neonatology unit was clean0.7330.404 The unit could easily be reached by telephone0.809 Our child’s incubator or bed was clean0.8 The team worked efficiently0.661 There was enough space around our child’s incubator/bed0.9 Noise in the unit was muffled as good as possible0.788\\nProfessional Attitude\\n Our child’s health always came first for the doctors and nurses0.799 The team worked hygienically0.747 Our cultural background was taken into account0.728 The doctors and nurses always took time to listen to us0.876 We felt welcome by the team0.804 Despite the workload, sufficient attention was paid to our child and us by the team0.5840.508 Nurses and doctors always introduced themselves by name and function0.794 We received sympathy from the doctors and nurses0.786 At our bedside, the discussion between the doctors and nurses was only about our child.0.4570.582 The team respected the privacy of our children and us.0.874 There was a pleasant atmosphere among the staff0.743 The team showed respect for our child and us0.5190.609\\n\\nFactor loadings and identified components’ of EMPATHIC-N tool\",\n \"The total EMPATHIC-N values and the five domains showed a good level of internal consistency. The five domains’ Cronbach’s values range from 0.639 to 0.791, whereas the total EMPATHIC-N Cronbach’s value was 0.904. The inter-item correlation (IIC) of the five domains had significant internal consistency (Table 3).\\n\\nTable 3Reliability of items of parental satisfaction with NICU-care serviceDimensionNumber of ItemsΧηρονβαχη αMean dimension score (SD)Maximum possible dimension scoreInter-item correlation (IIC)Item-discriminant validity (IDV)Information120.63939.55(8.42)720.22–0.48*0.20–0.57Care & Treatment170.79181.70(20.49)1700.14–0.64*0.25–0.38Parental Participation80.75524.86(7.79)480.01–0.71*0.37–0.54Organization80.72925.22(7.39)480.11–0.64*0.34–0.69Professional Attitude120.69438.106 (9.07)720.13–0.56*0.22–0.45EMPATHIC-N tool570.904209.44(42.82)410cc*Significant value (p < 0.001), c = not computable\\n\\nReliability of items of parental satisfaction with NICU-care service\\n*Significant value (p < 0.001), c = not computable\",\n \"Overall mean satisfaction level of parents in NICU-care service was 47.8% [95% CI= (43.1–52.5)]. The domains of parental satisfaction scores were compared with each other and, the lowest parental satisfaction score was for the domain of care and treatment 36.9% while the organizational domain showed the highest parental satisfaction level 59.0%. The domains of information, parental participation, and professional attitude showed comparable parental satisfaction scores (Fig. 1).\\n\\nFig. 1Parents’ overall and dimensional satisfaction with NICU-care services\\n\\nParents’ overall and dimensional satisfaction with NICU-care services\",\n \"In this study, the bivariable logistic regression showed that parental gender, residency, parental hospital stay, neonatal hospital stay, birth weight, and gestational age were factors with a p-value of less than 0.2 and were fitted with a multivariable logistic regression model. However, neonatal hospital stay was not associated with multivariable logistic regression with a p-value greater than 0.05. The multivariable logistic analyses showed that mothers were 2.16 (AOR = 2.16; 95%CI: 1.28–3.63) times more satisfied than fathers. Also, parents who are from the rural area were 2.94 (AOR = 2.94; 95%CI: 1.42–6.06) more satisfied than urban. Parents who say less than 15 days were 2.18 (AOR = 2.18; 95%CI: 1.13–4.20) times more satisfied than parents who stay 15 or more days in the hospital. Also, parents of a neonate with a normal birth weight of 2.14 (AOR = 2.14; 95%CI: 1.16–3.94) and full-term neonate 2.53(AOR = 2.53; 95%CI: 1.29–4.97) times more satisfied than their counterparts (Table 4).\\n\\nTable 4Factors associated with satisfaction of parents in NICU-care service (n = 385)VariablesSatisfaction level on NICU- serviceCrude odds ratioAdjusted odds ratiop-valueSatisfiedNot Satisfied(95% CI)(95% CI)\\nGender\\n Mother117(52.2%)107(47.8%)1.53(1.02,2.31)2.16(1.28,3.63)0.004* Father67(41.6%)94(58.4%)1\\nResidency\\n Urban68(36.4%)119(63.6%)1 Rural116(58.6%)82(41.4%)2.48(1.64,3.73)2.94(1.42,6.06)0.004*\\nParental hospital stay\\n ≤ 15 days140(57.4%)104(42.6%)2.97(1.92,4.59)2.18(1.13,4.20)0.019* > 15days44(31.2%)97(68.8%)1\\nBirth weight\\n High birth weight34(41.0%)49(59.0%)1 Normal birth weight96(59.3%)66(40.7%)2.09(1.22,3.59)2.14(1.16,3.94)0.015* Low birth weight54(38.6%)86(61.4%)0.91(0.52,1.58)0.76(0.41,1.41)0.38\\nGestational age\\n Premature18(30.0%)42(70.0%)1 Full term166(51.1%)159(48.9%)2.44(1.35,4.41)2.53(1.29,4.97)0.007**= p-value < 0.05 1 = reference BWt-birth weightNote: p-values were extracted from the multivariate logistic regression model\\n\\nFactors associated with satisfaction of parents in NICU-care service (n = 385)\\n*= p-value < 0.05 1 = reference BWt-birth weight\\nNote: p-values were extracted from the multivariate logistic regression model\",\n \"In this study, the total average parental satisfaction score with NICU service was 47.8%. This finding is nearly similar to a study done in Ethiopia (50%) [26]. Contrary to this finding, studies done in Norway (76%) [7] and Greece (99%) [27] had a higher parental satisfaction score with NICU service.\\nThe study found that parental satisfaction with NICU service was 50.4% in the information subscale, 36.9% in the care and treatment subscale, 50.1% in the parental participation subscale, 59.0% in the organization subscale, and 48.6% in the professional attitude subscale. Parental satisfaction was lowest in the care and treatment subscale. A study in Italy [20] and South Africa [28] on the other hand, discovered the lowest parental satisfaction score in the professional attitude and parental participation subscales respectively. This disparity could be attributed to the NICU’s lack of professional and medical resources to treat and care for neonates in this situation.\\nAccording to this study, mothers, parents from rural regions, parents of neonates who stayed less than 15 days in the hospital, parents of neonates with normal birth weight, and parents of full-term neonates were all more satisfied with NICU services than their counterparts.\\nThis study showed that mothers were more satisfied with NICU service than fathers. Similar studies conducted in Italy [20], Israel [11], and Greece [29] also showed that mothers were more satisfied than fathers in NICU service. The possible explanation might be women are allowed to spend more time in the NICU and participate in the care of their newborns and cultivate more relationships with medical caregivers than fathers.\\nAlso in this study parents who came from rural areas were more satisfied than parents from urban areas. This founding was also similar to a study done in Greece [30] found that parents from rural areas were more satisfied. The probable reason for this might be parents from rural areas may have a low-level awareness of the hospital, expectations, and demand for NICU service as compared with actual practice.\\nParents of normal birth weight and full-term neonates were more satisfied with NICU service than parents of low birth weight and preterm neonates in this study. In addition, a Norwegian study [7] found that parents of newborns with normal birth weights were more satisfied than those with low birth weights. Reasonably, parents of full-term and normal-weight neonates are likely to face unexpected concerns, which may lead to greater satisfaction with positive outcomes.\\nIn this study, parents of neonates who stayed less than 15 days in the hospital were more satisfied with NICU service than parents who stayed 15 or more days. Similarly, studies done in Ethiopia [26] and Iran [10] found parents who stayed for a short period in the NICU were more satisfied. The possible reason for this might be that parents with short-stay are less likely to see their neonatal serious conditions that make emotional and care mismanagements.\\n[SUBTITLE] Limitations of the study [SUBSECTION] The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.\\nThe study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.\",\n \"The study’s limitations include being a single-center nature and the lack of analysis of various alternatives to logistic regression that might be more applicable for this investigation.\",\n \"There was a low level of parental satisfaction with neonatal intensive care unit service. Among the dimensions of EMPATHIC-N, the lowest parental satisfaction score was in the care and treatment while the highest parental satisfaction score was in the organization dimension. As a result, health professionals and hospital administrators should collaborate to improve NICU services to provide high-quality service and satisfy parents.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction",null,null,null,null,null,null,null,null,null,"results",null,null,null,null,"discussion",null,"conclusion"],"string":"[\n \"introduction\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n \"conclusion\"\n]"},"keywords":{"kind":"list like","value":["NICU","Parents","Satisfaction","Service"],"string":"[\n \"NICU\",\n \"Parents\",\n \"Satisfaction\",\n \"Service\"\n]"}}},{"rowIdx":356,"cells":{"title":{"kind":"string","value":"'Corona is coming': COVID-19 vaccination perspectives and experiences amongst Culturally and Linguistically Diverse West Australians."},"pmid":{"kind":"string","value":"36262050"},"background_abstract":{"kind":"string","value":"Culturally and Linguistically Diverse (CALD) groups within high-income countries are at risk of being left behind by the COVID-19 vaccination rollout. They face both access and attitudinal barriers, including low trust in government and health authorities."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Perth, WA's capital, was chosen as the state's study site because most of the state's CALD population lives there. Eleven semistructured in-depth interviews and three focus groups (with 37 participants) were conducted with CALD residents between August and October 2021. Thematic analysis was conducted, informed by the 'Capability', 'Opportunity', 'Motivation', 'Behaviour' model."},"methods_abstract_label":{"kind":"string","value":"DESIGN AND PARTICIPANTS"},"results_abstract":{"kind":"string","value":"CALD participants faced barriers including a lack of knowledge about COVID-19 and the vaccines, low self-rated English proficiency and education levels, misinformation, passive government communication strategies and limited access to vaccine clinics/providers. They were, however, motivated to vaccinate by the imminent opening of state and international borders, trust in government and healthcare authorities, travel intentions and the desire to protect themselves and others."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Despite high levels of trust and significant desire for vaccines among CALD communities in Perth, current strategies were not meeting their needs and the community remains at risk from COVID-19. Tailored intervention strategies are required to provide knowledge, address misinformation and facilitate access to ensure uptake of COVID-19 vaccines-including for additional doses-amongst CALD communities. Governments should work with trusted CALD community members to disseminate tailored COVID-19 vaccine information and adequately translated resources."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Female","COVID-19 Vaccines","Australia","COVID-19","Cultural Diversity","Vaccination"],"string":"[\n \"Humans\",\n \"Female\",\n \"COVID-19 Vaccines\",\n \"Australia\",\n \"COVID-19\",\n \"Cultural Diversity\",\n \"Vaccination\"\n]"},"pmcid":{"kind":"string","value":"9700143"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Study design [SUBSECTION] This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\nThis CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\n[SUBTITLE] Consumer consultation process [SUBSECTION] Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\nConsumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\n[SUBTITLE] Study setting [SUBSECTION] In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.\nIn the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.\n[SUBTITLE] Sampling, recruitment and data collection [SUBSECTION] Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\nInterviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\n[SUBTITLE] Interview guide [SUBSECTION] Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.\nInterview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.\n[SUBTITLE] Analysis [SUBSECTION] All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\nAll interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set."},"results_title":{"kind":"null"},"results_text":{"kind":"null"},"conclusions_title":{"kind":"string","value":"CONCLUSION"},"conclusions_text":{"kind":"string","value":"This study provides in‐depth insight into the perceptions and attitudes of CALD communities in Perth, WA, regarding the COVID‐19 vaccine, identifying facilitators and barriers for engagement and uptake. CALD participants faced barriers including a lack of knowledge about COVID‐19 and the vaccines, low English proficiency and education levels, misinformation, passive government communication strategies and limited access to vaccine clinics/providers. They were, however, motivated to vaccinate by the imminent opening of state and international borders, trust in government and healthcare authorities, travel intentions and the desire to protect themselves and others. Despite high levels of trust and significant desire for vaccines, strategies at the time of our study were not meeting the CALD community's needs, and thus those within it remain at risk from the disease. Tailored intervention strategies are required to provide knowledge, address misinformation, and facilitate access to ensure uptake of COVID‐19 vaccines—including for additional doses—amongst CALD communities. Governments must work with trusted CALD community members and immunization providers such as GPs to disseminate tailored COVID‐19 vaccine information and adequately translated resources so that community members themselves can become stronger vaccine ambassadors to each other."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Study design","Consumer consultation process","Study setting","Sampling, recruitment and data collection","Interview guide","Analysis","FINDINGS","Capability","Knowledge about COVID‐19 and COVID‐19 vaccination","Knowledge about how and where to get vaccinated","Opportunity","Relationships with general practitioners (GPs)","Access to information/misinformation","Access to vaccine clinics","Community influence","Motivation","IMPLICATIONS FOR POLICY AND PRACTICE","Limitations","AUTHOR CONTRIBUTIONS"],"string":"[\n \"INTRODUCTION\",\n \"Study design\",\n \"Consumer consultation process\",\n \"Study setting\",\n \"Sampling, recruitment and data collection\",\n \"Interview guide\",\n \"Analysis\",\n \"FINDINGS\",\n \"Capability\",\n \"Knowledge about COVID‐19 and COVID‐19 vaccination\",\n \"Knowledge about how and where to get vaccinated\",\n \"Opportunity\",\n \"Relationships with general practitioners (GPs)\",\n \"Access to information/misinformation\",\n \"Access to vaccine clinics\",\n \"Community influence\",\n \"Motivation\",\n \"IMPLICATIONS FOR POLICY AND PRACTICE\",\n \"Limitations\",\n \"AUTHOR CONTRIBUTIONS\"\n]"},"other_sections_texts":{"kind":"list like","value":["Scholars and commentators recognize that even within high‐income countries, pockets of populations are at risk of being left behind in the globally inequitable COVID‐19 vaccine rollout.\n1\n Culturally and Linguistically Diverse (CALD) populations, which include refugees and migrants, are at particular risk due to inaccessible or suboptimal healthcare experiences, which may compound attitudinal factors, including vaccine hesitancy.\n2\n, \n3\n Unique local factors can also create barriers: in the case of Western Australia (WA), almost nonexistent levels of community transmission of COVID‐19 led some locals to ‘wait awhile’ before being vaccinated as state borders remained closed throughout 2021.\n4\n Meanwhile, earlier large outbreaks in other Australian states disproportionately affected CALD communities, with significantly higher morbidity and mortality.\n5\n, \n6\n\n\nIn Australia, COVID‐19 vaccination is recorded on the Australian Immunisation Register, however, country of birth and language spoken at home are not captured on this register. Therefore, there is no publicly available data on uptake among CALD communities in Australia. However, an online survey of refugees in Australia in mid‐2021 identified that 88% had not yet received a COVID‐19 vaccine.\n7\n Emerging evidence suggests that COVID‐19 vaccine hesitancy is prevalent among Australian CALD populations.\n8\n, \n9\n Focus groups conducted in mid‐2021 in New South Wales with CALD people indicated that 42% of participants were not planning to be vaccinated, and 29% were unsure or hesitant.\n8\n Many had safety concerns, were uncertain how vaccines work or lacked trust in government.\n8\n These uncertainties are unsurprising, given little of the public health information on COVID‐19 in Australia was tailored towards those with low English proficiency or health literacy.\n10\n Lockyer et al.'s\n11\n qualitative study in multicultural Bradford, UK, also found that participants' trust in government and the National Health Service dropped during the pandemic and that vaccine hesitancy was expressed by ethnic or national minorities.\nIn this study, we sought to understand the perceptions and attitudes to COVID‐19 vaccines amongst WA's CALD communities, to identify barriers preventing uptake, and to consider strategies to promote uptake in this population. Findings can support the development of strategies to enhance uptake among the CALD communities locally, nationwide and in comparable high‐income English‐speaking countries. They are particularly pertinent in the context of needing to enhance third and fourth‐dose rates for adults and protecting CALD children with paediatric COVID‐19 vaccinations.","This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).","Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.","In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.","Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.","Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.","All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.","Drawing on data from our interviews and focus groups, we created 10 subthemes encompassed within the three superordinate themes of Capability, Opportunity and Motivation (Figure 1).\nThe Capability, Opportunity and Motivation influencing COVID‐19 vaccine uptake among Culturally and Linguistically Diverse individuals in Perth, WA. Figure modified from COM‐B model\n24\n; arrows indicate influence of factor on other factors.\n[SUBTITLE] Capability [SUBSECTION] [SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n[SUBTITLE] Opportunity [SUBSECTION] [SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Motivation [SUBSECTION] Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\nMost participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)","[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)","Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])","Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)","[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)","Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)","Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.","Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.","Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)","Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)","A strong motivation for this research was to assist in strategic design, informing ongoing and future implementation of public health initiatives among CALD populations. Our experiences engaging with our CALD participants and their partners, including helping them to book appointments after the focus groups, emphasized how important it is that governments promptly address vaccine access issues, as well as campaigning and working directly with CALD groups, and featuring them visibly in messaging. The barriers regarding lack of access to resources and the circulation of misinformation must be addressed as soon as possible to reduce the COVID‐19 burden amongst the CALD population in WA. These lessons can then be utilized for future vaccine doses, for reaching the children of CALD populations, and for wider public health initiatives.\nAt the time of our study, CALD participants and their communities were not publicly sharing their vaccinated status and proudly promoting COVID‐19 vaccinations. However, it is unreasonable to demand more social courage from CALD individuals when what they need is more support. Given that governments are responsible for making vaccination accessible and acceptable to improve and sustain uptake,\n34\n they must recognize, work with, and train influential community leaders on the importance of and how to have meaningful vaccine conversations with their peers.\n32\n A recent review identified the who, the what and the how for effective community engagement during a pandemic: essentially, it is important to engage with the community from the very beginning for a better chance of behaviour change within the community.\n35\n In eastern Australia, Mahimbo et al.\n9\n identified that CALD communities are eager for community engagement, and for their leaders to champion the vaccine. Seale et al.\n29\n identified that while some CALD organizations in Australia were updating local CALD leaders on relevant COVID‐19 information to share with their community, there was a lack of training in how to do so. Training influential community leaders can occur over a relatively short period of time: a COVID‐19 information and support programme was offered in the United States to CALD community leaders which involved 1‐hour conference calls twice a week (we acknowledge that this is then reliant on the leaders having a decent level of computer literacy.)\n29\n Further, given the trust and respect our participants held for their health practitioners, it is imperative that governments support GPs to effectively communicate about COVID‐19 vaccination with CALD patients; this requires resources, training and remuneration.\n36\n Since our data collection, the Australian National Centre for Immunisation Research and Surveillance have published a COVID‐19 decision aid in Greek, Vietnamese, Arabic, Simplified Chinese and Traditional Chinese.\n37\n Immunisation providers could use or share this tool with their CALD patients (though its effectiveness in increasing vaccine uptake is yet to be evaluated).\nOur participants' high levels of trust in governing authorities is an untapped opportunity. WA participants knew that ‘Corona is coming’. Now that it has arrived, the government and healthcare providers must maintain the trust of CALD by reaching them promptly with vaccine doses and culturally appropriate information to reduce the burden of COVID‐19. If these measures fail, it would be entirely reasonable for CALD individuals to not blame their own choices, and instead look to the authorities they believed would protect them and ask why they were not offered greater support and access.\n[SUBTITLE] Limitations [SUBSECTION] Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\nQualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.","Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.","Samantha J. Carlson supervised the data collection and analysis, led the initial manuscript drafting, contributed to the study design, contributed to project administration and coordinated the ethics and stakeholder engagement. Gracie Edwards collected the data and conducted the analysis as well as drafted the literature review and contributed to the overall manuscript drafting. Christopher C. Blyth contributed to the conceptualization of the project and provided oversight and input into the study and funding attainment, and reviewed and edited the manuscript. Barbara Nattabi supervised the data collection and analysis, contributed to the coding framework, provided CALD expertise, contributed to project administration and reviewed and edited the manuscript. Katie Attwell conceptualized the project, co‐developed the broader study methodology with Christopher C. Blyth, led the funding attainment, and supervised the conduct and administration of the broader research project. She contributed to the literature review and the drafting, reviewing and editing of the manuscript. All authors have approved the final manuscript."],"string":"[\n \"Scholars and commentators recognize that even within high‐income countries, pockets of populations are at risk of being left behind in the globally inequitable COVID‐19 vaccine rollout.\\n1\\n Culturally and Linguistically Diverse (CALD) populations, which include refugees and migrants, are at particular risk due to inaccessible or suboptimal healthcare experiences, which may compound attitudinal factors, including vaccine hesitancy.\\n2\\n, \\n3\\n Unique local factors can also create barriers: in the case of Western Australia (WA), almost nonexistent levels of community transmission of COVID‐19 led some locals to ‘wait awhile’ before being vaccinated as state borders remained closed throughout 2021.\\n4\\n Meanwhile, earlier large outbreaks in other Australian states disproportionately affected CALD communities, with significantly higher morbidity and mortality.\\n5\\n, \\n6\\n\\n\\nIn Australia, COVID‐19 vaccination is recorded on the Australian Immunisation Register, however, country of birth and language spoken at home are not captured on this register. Therefore, there is no publicly available data on uptake among CALD communities in Australia. However, an online survey of refugees in Australia in mid‐2021 identified that 88% had not yet received a COVID‐19 vaccine.\\n7\\n Emerging evidence suggests that COVID‐19 vaccine hesitancy is prevalent among Australian CALD populations.\\n8\\n, \\n9\\n Focus groups conducted in mid‐2021 in New South Wales with CALD people indicated that 42% of participants were not planning to be vaccinated, and 29% were unsure or hesitant.\\n8\\n Many had safety concerns, were uncertain how vaccines work or lacked trust in government.\\n8\\n These uncertainties are unsurprising, given little of the public health information on COVID‐19 in Australia was tailored towards those with low English proficiency or health literacy.\\n10\\n Lockyer et al.'s\\n11\\n qualitative study in multicultural Bradford, UK, also found that participants' trust in government and the National Health Service dropped during the pandemic and that vaccine hesitancy was expressed by ethnic or national minorities.\\nIn this study, we sought to understand the perceptions and attitudes to COVID‐19 vaccines amongst WA's CALD communities, to identify barriers preventing uptake, and to consider strategies to promote uptake in this population. Findings can support the development of strategies to enhance uptake among the CALD communities locally, nationwide and in comparable high‐income English‐speaking countries. They are particularly pertinent in the context of needing to enhance third and fourth‐dose rates for adults and protecting CALD children with paediatric COVID‐19 vaccinations.\",\n \"This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\\n12\\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\",\n \"Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\",\n \"In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\\n13\\n with transmission rapidly increasing only from February 2022.\\n13\\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\\n14\\n and over 85.5% of those eligible for their third dose in WA having received it.\\n15\\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\\n16\\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\\n17\\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\\n18\\n although at the time of our study only approximately 20 were in place.\\nCALD individuals comprise almost one‐fifth of the WA population.\\n19\\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\\n19\\n so we focused on recruiting CALD adults living in Perth.\",\n \"Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\\n20\\n, \\n21\\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\\n22\\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\\n19\\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\\nSome participants spoke ≥1 language.\\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\",\n \"Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\\n12\\n The questions were simplified for the focus groups.\",\n \"All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\\n23\\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\\n24\\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\\n25\\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\",\n \"Drawing on data from our interviews and focus groups, we created 10 subthemes encompassed within the three superordinate themes of Capability, Opportunity and Motivation (Figure 1).\\nThe Capability, Opportunity and Motivation influencing COVID‐19 vaccine uptake among Culturally and Linguistically Diverse individuals in Perth, WA. Figure modified from COM‐B model\\n24\\n; arrows indicate influence of factor on other factors.\\n[SUBTITLE] Capability [SUBSECTION] [SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n[SUBTITLE] Opportunity [SUBSECTION] [SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Motivation [SUBSECTION] Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\nMost participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\",\n \"[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\",\n \"Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\",\n \"Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\",\n \"[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\",\n \"Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\",\n \"Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\",\n \"A strong motivation for this research was to assist in strategic design, informing ongoing and future implementation of public health initiatives among CALD populations. Our experiences engaging with our CALD participants and their partners, including helping them to book appointments after the focus groups, emphasized how important it is that governments promptly address vaccine access issues, as well as campaigning and working directly with CALD groups, and featuring them visibly in messaging. The barriers regarding lack of access to resources and the circulation of misinformation must be addressed as soon as possible to reduce the COVID‐19 burden amongst the CALD population in WA. These lessons can then be utilized for future vaccine doses, for reaching the children of CALD populations, and for wider public health initiatives.\\nAt the time of our study, CALD participants and their communities were not publicly sharing their vaccinated status and proudly promoting COVID‐19 vaccinations. However, it is unreasonable to demand more social courage from CALD individuals when what they need is more support. Given that governments are responsible for making vaccination accessible and acceptable to improve and sustain uptake,\\n34\\n they must recognize, work with, and train influential community leaders on the importance of and how to have meaningful vaccine conversations with their peers.\\n32\\n A recent review identified the who, the what and the how for effective community engagement during a pandemic: essentially, it is important to engage with the community from the very beginning for a better chance of behaviour change within the community.\\n35\\n In eastern Australia, Mahimbo et al.\\n9\\n identified that CALD communities are eager for community engagement, and for their leaders to champion the vaccine. Seale et al.\\n29\\n identified that while some CALD organizations in Australia were updating local CALD leaders on relevant COVID‐19 information to share with their community, there was a lack of training in how to do so. Training influential community leaders can occur over a relatively short period of time: a COVID‐19 information and support programme was offered in the United States to CALD community leaders which involved 1‐hour conference calls twice a week (we acknowledge that this is then reliant on the leaders having a decent level of computer literacy.)\\n29\\n Further, given the trust and respect our participants held for their health practitioners, it is imperative that governments support GPs to effectively communicate about COVID‐19 vaccination with CALD patients; this requires resources, training and remuneration.\\n36\\n Since our data collection, the Australian National Centre for Immunisation Research and Surveillance have published a COVID‐19 decision aid in Greek, Vietnamese, Arabic, Simplified Chinese and Traditional Chinese.\\n37\\n Immunisation providers could use or share this tool with their CALD patients (though its effectiveness in increasing vaccine uptake is yet to be evaluated).\\nOur participants' high levels of trust in governing authorities is an untapped opportunity. WA participants knew that ‘Corona is coming’. Now that it has arrived, the government and healthcare providers must maintain the trust of CALD by reaching them promptly with vaccine doses and culturally appropriate information to reduce the burden of COVID‐19. If these measures fail, it would be entirely reasonable for CALD individuals to not blame their own choices, and instead look to the authorities they believed would protect them and ask why they were not offered greater support and access.\\n[SUBTITLE] Limitations [SUBSECTION] Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\\nQualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\",\n \"Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\",\n \"Samantha J. Carlson supervised the data collection and analysis, led the initial manuscript drafting, contributed to the study design, contributed to project administration and coordinated the ethics and stakeholder engagement. Gracie Edwards collected the data and conducted the analysis as well as drafted the literature review and contributed to the overall manuscript drafting. Christopher C. Blyth contributed to the conceptualization of the project and provided oversight and input into the study and funding attainment, and reviewed and edited the manuscript. Barbara Nattabi supervised the data collection and analysis, contributed to the coding framework, provided CALD expertise, contributed to project administration and reviewed and edited the manuscript. Katie Attwell conceptualized the project, co‐developed the broader study methodology with Christopher C. Blyth, led the funding attainment, and supervised the conduct and administration of the broader research project. She contributed to the literature review and the drafting, reviewing and editing of the manuscript. All authors have approved the final manuscript.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Study design","Consumer consultation process","Study setting","Sampling, recruitment and data collection","Interview guide","Analysis","FINDINGS","Capability","Knowledge about COVID‐19 and COVID‐19 vaccination","Knowledge about how and where to get vaccinated","Opportunity","Relationships with general practitioners (GPs)","Access to information/misinformation","Access to vaccine clinics","Community influence","Motivation","DISCUSSION","IMPLICATIONS FOR POLICY AND PRACTICE","Limitations","CONCLUSION","AUTHOR CONTRIBUTIONS","CONFLICT OF INTEREST"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Study design\",\n \"Consumer consultation process\",\n \"Study setting\",\n \"Sampling, recruitment and data collection\",\n \"Interview guide\",\n \"Analysis\",\n \"FINDINGS\",\n \"Capability\",\n \"Knowledge about COVID‐19 and COVID‐19 vaccination\",\n \"Knowledge about how and where to get vaccinated\",\n \"Opportunity\",\n \"Relationships with general practitioners (GPs)\",\n \"Access to information/misinformation\",\n \"Access to vaccine clinics\",\n \"Community influence\",\n \"Motivation\",\n \"DISCUSSION\",\n \"IMPLICATIONS FOR POLICY AND PRACTICE\",\n \"Limitations\",\n \"CONCLUSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"CONFLICT OF INTEREST\"\n]"},"all_sections_texts":{"kind":"list like","value":["Scholars and commentators recognize that even within high‐income countries, pockets of populations are at risk of being left behind in the globally inequitable COVID‐19 vaccine rollout.\n1\n Culturally and Linguistically Diverse (CALD) populations, which include refugees and migrants, are at particular risk due to inaccessible or suboptimal healthcare experiences, which may compound attitudinal factors, including vaccine hesitancy.\n2\n, \n3\n Unique local factors can also create barriers: in the case of Western Australia (WA), almost nonexistent levels of community transmission of COVID‐19 led some locals to ‘wait awhile’ before being vaccinated as state borders remained closed throughout 2021.\n4\n Meanwhile, earlier large outbreaks in other Australian states disproportionately affected CALD communities, with significantly higher morbidity and mortality.\n5\n, \n6\n\n\nIn Australia, COVID‐19 vaccination is recorded on the Australian Immunisation Register, however, country of birth and language spoken at home are not captured on this register. Therefore, there is no publicly available data on uptake among CALD communities in Australia. However, an online survey of refugees in Australia in mid‐2021 identified that 88% had not yet received a COVID‐19 vaccine.\n7\n Emerging evidence suggests that COVID‐19 vaccine hesitancy is prevalent among Australian CALD populations.\n8\n, \n9\n Focus groups conducted in mid‐2021 in New South Wales with CALD people indicated that 42% of participants were not planning to be vaccinated, and 29% were unsure or hesitant.\n8\n Many had safety concerns, were uncertain how vaccines work or lacked trust in government.\n8\n These uncertainties are unsurprising, given little of the public health information on COVID‐19 in Australia was tailored towards those with low English proficiency or health literacy.\n10\n Lockyer et al.'s\n11\n qualitative study in multicultural Bradford, UK, also found that participants' trust in government and the National Health Service dropped during the pandemic and that vaccine hesitancy was expressed by ethnic or national minorities.\nIn this study, we sought to understand the perceptions and attitudes to COVID‐19 vaccines amongst WA's CALD communities, to identify barriers preventing uptake, and to consider strategies to promote uptake in this population. Findings can support the development of strategies to enhance uptake among the CALD communities locally, nationwide and in comparable high‐income English‐speaking countries. They are particularly pertinent in the context of needing to enhance third and fourth‐dose rates for adults and protecting CALD children with paediatric COVID‐19 vaccinations.","[SUBTITLE] Study design [SUBSECTION] This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\nThis CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\n[SUBTITLE] Consumer consultation process [SUBSECTION] Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\nConsumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\n[SUBTITLE] Study setting [SUBSECTION] In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.\nIn the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.\n[SUBTITLE] Sampling, recruitment and data collection [SUBSECTION] Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\nInterviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\n[SUBTITLE] Interview guide [SUBSECTION] Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.\nInterview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.\n[SUBTITLE] Analysis [SUBSECTION] All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\nAll interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.","This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\n12\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).","Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.","In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\n13\n with transmission rapidly increasing only from February 2022.\n13\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\n14\n and over 85.5% of those eligible for their third dose in WA having received it.\n15\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\n16\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\n17\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\n18\n although at the time of our study only approximately 20 were in place.\nCALD individuals comprise almost one‐fifth of the WA population.\n19\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\n19\n so we focused on recruiting CALD adults living in Perth.","Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\n20\n, \n21\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\n22\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\n19\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\nSome participants spoke ≥1 language.\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.","Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\n12\n The questions were simplified for the focus groups.","All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\n23\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\n24\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\n25\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.","Drawing on data from our interviews and focus groups, we created 10 subthemes encompassed within the three superordinate themes of Capability, Opportunity and Motivation (Figure 1).\nThe Capability, Opportunity and Motivation influencing COVID‐19 vaccine uptake among Culturally and Linguistically Diverse individuals in Perth, WA. Figure modified from COM‐B model\n24\n; arrows indicate influence of factor on other factors.\n[SUBTITLE] Capability [SUBSECTION] [SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n[SUBTITLE] Opportunity [SUBSECTION] [SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Motivation [SUBSECTION] Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\nMost participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)","[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)","Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\n\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])","Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\n\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)","[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)","Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\n\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)","Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\n\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\n\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\n\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\n\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\n\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\n\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\n\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.","Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\n\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\n\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.","Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\n\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\n\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\n\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\n\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\n\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\n\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\n\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)","Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\n\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\n\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)","This is the first study to explore perceptions and attitudes to the COVID‐19 vaccine, barriers to preventing and strategies promoting vaccine uptake among CALD populations in Perth, WA. Similar to studies from the United Kingdom\n11\n, \n26\n and New South Wales,\n8\n we found high levels of COVID‐19 vaccine hesitancy among CALD individuals due to safety concerns, lack of understanding and high levels of misinformation perpetuated by CALD peers or social media platforms. Hesitancy and low uptake were particularly apparent among participants with low levels of education and low English proficiency. These findings have since been confirmed in a large regression analysis of over 12 million Australians, which identified that those with low levels of education and English proficiency were less likely to have received their first COVID‐19 vaccine dose.\n27\n\n\nDespite their concerns, participants were not completely opposed to vaccination, and we were struck by how they did not see the COVID‐19 vaccination programme as being for people like them. They were exposed to misinformation through their social networks, as previous studies have also found,\n11\n and the hesitancy this generated is a wider problem. However, participants also felt that governments were not forthcoming with resources to address their concerns. Participants were ostensibly included in cohorts that were eligible for vaccination at the time of data collection, and some of them had been eligible many months prior, due to their age. However, although they were ostensibly being asked to ‘roll up’, they received few or no personal cues to avail themselves of vaccinations.\nPractical barriers were also evident. As previously identified by Abdi et al.,\n28\n refugees and migrants tend to have quite demanding competing priorities, such as seeking employment and accommodation and thus need prompting for vaccination. In some cases, our participants' efforts to be vaccinated were thwarted by the online booking system. In fact, some of our research team assisted participants to book vaccination appointments on their smartphones after the focus groups or provide further information to partners who waited in the car park to speak with them. It is not just our participants who struggled with the online booking system: Seale et al.\n29\n describe a ‘digital divide’ in which those without access to the internet and have low English proficiency are disadvantaged in their ability to access COVID‐19 vaccines, and Biddle et al.\n27\n identified that Australians who completed the 2016 census on paper rather than online had a lower probability of vaccination, suggesting that individuals with lower digital literacy may experience more barriers to vaccination. Booking systems must be accessible to varying levels of digital literacy and English proficiency.\nSocial norms are a strong influence on vaccine uptake.\n16\n, \n30\n For CALD people in Perth, social norms were blocking COVID‐19 vaccination. Vaccinated CALD individuals were not sharing their experiences or reasons for vaccinating, facilitating ongoing shame and stigma, and thereby perpetuating continued silence for those who did vaccinate; tendencies that have also been found in local ‘alternative’ communities with regard to childhood vaccination.\n31\n Some CALD leaders were also not actively encouraging vaccination due to their own concerns or opposition, an issue identified early on by the Australian multicultural sector as a potential barrier.\n32\n\n\nAnother well‐established influence on vaccine uptake is receiving a recommendation from a healthcare provider.\n30\n Our participants showed great trust in their GPs and reported positive experiences of consultation. Accordingly, they believed that GPs would have more opportunities to address COVID‐19 vaccine concerns than mass vaccination clinic staff. However, GPs may not always be easily accessible within an acceptable timeframe to the patient,\n33\n meaning some CALD individuals are turning to or relying on social media for their health/COVID‐19 information.\n9\n\n\nOur participants' high levels of trust in governing authorities was a novel finding, mostly related to feeling safe from COVID‐19 in WA. Focus groups with CALD individuals in New South Wales instead found participants were concerned that the government was using COVID‐19 vaccination to control people.\n8\n We were struck by how our participants' trust in government emanated from people whose experiences in their countries of origin had often been traumatic, culminating in their refugee status.","A strong motivation for this research was to assist in strategic design, informing ongoing and future implementation of public health initiatives among CALD populations. Our experiences engaging with our CALD participants and their partners, including helping them to book appointments after the focus groups, emphasized how important it is that governments promptly address vaccine access issues, as well as campaigning and working directly with CALD groups, and featuring them visibly in messaging. The barriers regarding lack of access to resources and the circulation of misinformation must be addressed as soon as possible to reduce the COVID‐19 burden amongst the CALD population in WA. These lessons can then be utilized for future vaccine doses, for reaching the children of CALD populations, and for wider public health initiatives.\nAt the time of our study, CALD participants and their communities were not publicly sharing their vaccinated status and proudly promoting COVID‐19 vaccinations. However, it is unreasonable to demand more social courage from CALD individuals when what they need is more support. Given that governments are responsible for making vaccination accessible and acceptable to improve and sustain uptake,\n34\n they must recognize, work with, and train influential community leaders on the importance of and how to have meaningful vaccine conversations with their peers.\n32\n A recent review identified the who, the what and the how for effective community engagement during a pandemic: essentially, it is important to engage with the community from the very beginning for a better chance of behaviour change within the community.\n35\n In eastern Australia, Mahimbo et al.\n9\n identified that CALD communities are eager for community engagement, and for their leaders to champion the vaccine. Seale et al.\n29\n identified that while some CALD organizations in Australia were updating local CALD leaders on relevant COVID‐19 information to share with their community, there was a lack of training in how to do so. Training influential community leaders can occur over a relatively short period of time: a COVID‐19 information and support programme was offered in the United States to CALD community leaders which involved 1‐hour conference calls twice a week (we acknowledge that this is then reliant on the leaders having a decent level of computer literacy.)\n29\n Further, given the trust and respect our participants held for their health practitioners, it is imperative that governments support GPs to effectively communicate about COVID‐19 vaccination with CALD patients; this requires resources, training and remuneration.\n36\n Since our data collection, the Australian National Centre for Immunisation Research and Surveillance have published a COVID‐19 decision aid in Greek, Vietnamese, Arabic, Simplified Chinese and Traditional Chinese.\n37\n Immunisation providers could use or share this tool with their CALD patients (though its effectiveness in increasing vaccine uptake is yet to be evaluated).\nOur participants' high levels of trust in governing authorities is an untapped opportunity. WA participants knew that ‘Corona is coming’. Now that it has arrived, the government and healthcare providers must maintain the trust of CALD by reaching them promptly with vaccine doses and culturally appropriate information to reduce the burden of COVID‐19. If these measures fail, it would be entirely reasonable for CALD individuals to not blame their own choices, and instead look to the authorities they believed would protect them and ask why they were not offered greater support and access.\n[SUBTITLE] Limitations [SUBSECTION] Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\nQualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.","Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.","This study provides in‐depth insight into the perceptions and attitudes of CALD communities in Perth, WA, regarding the COVID‐19 vaccine, identifying facilitators and barriers for engagement and uptake. CALD participants faced barriers including a lack of knowledge about COVID‐19 and the vaccines, low English proficiency and education levels, misinformation, passive government communication strategies and limited access to vaccine clinics/providers. They were, however, motivated to vaccinate by the imminent opening of state and international borders, trust in government and healthcare authorities, travel intentions and the desire to protect themselves and others. Despite high levels of trust and significant desire for vaccines, strategies at the time of our study were not meeting the CALD community's needs, and thus those within it remain at risk from the disease. Tailored intervention strategies are required to provide knowledge, address misinformation, and facilitate access to ensure uptake of COVID‐19 vaccines—including for additional doses—amongst CALD communities. Governments must work with trusted CALD community members and immunization providers such as GPs to disseminate tailored COVID‐19 vaccine information and adequately translated resources so that community members themselves can become stronger vaccine ambassadors to each other.","Samantha J. Carlson supervised the data collection and analysis, led the initial manuscript drafting, contributed to the study design, contributed to project administration and coordinated the ethics and stakeholder engagement. Gracie Edwards collected the data and conducted the analysis as well as drafted the literature review and contributed to the overall manuscript drafting. Christopher C. Blyth contributed to the conceptualization of the project and provided oversight and input into the study and funding attainment, and reviewed and edited the manuscript. Barbara Nattabi supervised the data collection and analysis, contributed to the coding framework, provided CALD expertise, contributed to project administration and reviewed and edited the manuscript. Katie Attwell conceptualized the project, co‐developed the broader study methodology with Christopher C. Blyth, led the funding attainment, and supervised the conduct and administration of the broader research project. She contributed to the literature review and the drafting, reviewing and editing of the manuscript. All authors have approved the final manuscript.","K. A. and C. C. B. are specialist advisors to the Australian Technical Advisory Group on Immunisation. The remaining authors declare no conflict of interest."],"string":"[\n \"Scholars and commentators recognize that even within high‐income countries, pockets of populations are at risk of being left behind in the globally inequitable COVID‐19 vaccine rollout.\\n1\\n Culturally and Linguistically Diverse (CALD) populations, which include refugees and migrants, are at particular risk due to inaccessible or suboptimal healthcare experiences, which may compound attitudinal factors, including vaccine hesitancy.\\n2\\n, \\n3\\n Unique local factors can also create barriers: in the case of Western Australia (WA), almost nonexistent levels of community transmission of COVID‐19 led some locals to ‘wait awhile’ before being vaccinated as state borders remained closed throughout 2021.\\n4\\n Meanwhile, earlier large outbreaks in other Australian states disproportionately affected CALD communities, with significantly higher morbidity and mortality.\\n5\\n, \\n6\\n\\n\\nIn Australia, COVID‐19 vaccination is recorded on the Australian Immunisation Register, however, country of birth and language spoken at home are not captured on this register. Therefore, there is no publicly available data on uptake among CALD communities in Australia. However, an online survey of refugees in Australia in mid‐2021 identified that 88% had not yet received a COVID‐19 vaccine.\\n7\\n Emerging evidence suggests that COVID‐19 vaccine hesitancy is prevalent among Australian CALD populations.\\n8\\n, \\n9\\n Focus groups conducted in mid‐2021 in New South Wales with CALD people indicated that 42% of participants were not planning to be vaccinated, and 29% were unsure or hesitant.\\n8\\n Many had safety concerns, were uncertain how vaccines work or lacked trust in government.\\n8\\n These uncertainties are unsurprising, given little of the public health information on COVID‐19 in Australia was tailored towards those with low English proficiency or health literacy.\\n10\\n Lockyer et al.'s\\n11\\n qualitative study in multicultural Bradford, UK, also found that participants' trust in government and the National Health Service dropped during the pandemic and that vaccine hesitancy was expressed by ethnic or national minorities.\\nIn this study, we sought to understand the perceptions and attitudes to COVID‐19 vaccines amongst WA's CALD communities, to identify barriers preventing uptake, and to consider strategies to promote uptake in this population. Findings can support the development of strategies to enhance uptake among the CALD communities locally, nationwide and in comparable high‐income English‐speaking countries. They are particularly pertinent in the context of needing to enhance third and fourth‐dose rates for adults and protecting CALD children with paediatric COVID‐19 vaccinations.\",\n \"[SUBTITLE] Study design [SUBSECTION] This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\\n12\\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\\nThis CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\\n12\\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\\n[SUBTITLE] Consumer consultation process [SUBSECTION] Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\\nConsumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\\n[SUBTITLE] Study setting [SUBSECTION] In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\\n13\\n with transmission rapidly increasing only from February 2022.\\n13\\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\\n14\\n and over 85.5% of those eligible for their third dose in WA having received it.\\n15\\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\\n16\\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\\n17\\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\\n18\\n although at the time of our study only approximately 20 were in place.\\nCALD individuals comprise almost one‐fifth of the WA population.\\n19\\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\\n19\\n so we focused on recruiting CALD adults living in Perth.\\nIn the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\\n13\\n with transmission rapidly increasing only from February 2022.\\n13\\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\\n14\\n and over 85.5% of those eligible for their third dose in WA having received it.\\n15\\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\\n16\\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\\n17\\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\\n18\\n although at the time of our study only approximately 20 were in place.\\nCALD individuals comprise almost one‐fifth of the WA population.\\n19\\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\\n19\\n so we focused on recruiting CALD adults living in Perth.\\n[SUBTITLE] Sampling, recruitment and data collection [SUBSECTION] Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\\n20\\n, \\n21\\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\\n22\\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\\n19\\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\\nSome participants spoke ≥1 language.\\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\\nInterviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\\n20\\n, \\n21\\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\\n22\\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\\n19\\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\\nSome participants spoke ≥1 language.\\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\\n[SUBTITLE] Interview guide [SUBSECTION] Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\\n12\\n The questions were simplified for the focus groups.\\nInterview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\\n12\\n The questions were simplified for the focus groups.\\n[SUBTITLE] Analysis [SUBSECTION] All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\\n23\\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\\n24\\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\\n25\\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\\nAll interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\\n23\\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\\n24\\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\\n25\\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\",\n \"This CALD study was conducted as part of the larger mixed methods research project, Coronavax. Coronavax was initiated in April 2020 with the aim of ascertaining community attitudes towards vaccination and vaccine access in WA through a series of qualitative interviews with segmented population groups. A key aspect of the project involved translating findings to the state and federal governments in real‐time via Functional Dialogues, a novel formal research exchange mechanism in which data on community attitudes are presented to government officials, who then contribute their own reflections and experiences as additional research data (not reported here).\\n12\\n Coronavax received ethics approval from the Child and Adolescent Services (CAHS) Human Research Ethics Committee (RGS0000004457).\",\n \"Consumer reference for Coronavax is led by Ms. Catherine Hughes who also chairs the Wesfarmers Centre of Vaccines and Infectious Diseases Community Reference Group at Telethon Kids Institute. This Group consulted on Coronavax in September 2020. Consumer involvement in Coronavax has sought to ensure the appropriateness and sensitivity of research questions and recruitment and to ensure that the research priorities and design reflect the needs of the community. Ishar Multicultural Women's Health Services, one of the largest multicultural health providers in Western Australia, provided additional consultation for this CALD study, as well as aiding with the recruitment of their clients, providing facilities, and hosting the focus groups.\",\n \"In the first 2 years of the COVID‐19 pandemic, WA experienced just over 1100 cases in a population of approximately 2.6 million people,\\n13\\n with transmission rapidly increasing only from February 2022.\\n13\\n As of March 2022, over 95% of people aged 12 years and older in WA had received at least two COVID‐19 vaccine doses,\\n14\\n and over 85.5% of those eligible for their third dose in WA having received it.\\n15\\n However, coverage rates at the time of our study were lower, as Australia had faced significant supply problems early in the rollout.\\n16\\n Responsibility for vaccine operations sat with both State and Federal Governments. WA residents access COVID‐19 vaccines through general practice clinics, federal respiratory clinics, pharmacies, community health services and state‐led mass vaccination clinics.\\nAs governments sought to rapidly increase vaccine uptake across the entire population, populations requiring more complex and bespoke interventions, such as CALD communities, remained under‐served. For example, Local Government Areas (LGAs) in Perth with the highest proportion of those born in nonmain English‐speaking countries had some of the lowest uptake in Perth 8 months into the rollout.\\n17\\n A more tailored approach to CALD communities commenced in late 2021 after our study, including pop‐up vaccination clinics at major religious events (e.g., the Indian festival of Diwali, November 2021), or at multicultural community centres. The state‐wide campaign ‘Roll Up For WA’ translated resources into over 50 non‐English languages,\\n18\\n although at the time of our study only approximately 20 were in place.\\nCALD individuals comprise almost one‐fifth of the WA population.\\n19\\n In WA, 90% of individuals born in what the Australian Bureau of Statistics calls ‘non‐main English‐speaking countries’ reside in the Perth metropolitan region,\\n19\\n so we focused on recruiting CALD adults living in Perth.\",\n \"Interviewees were recruited through CALD organizations, social media (including a specific campaign directed towards CALD men), and promotional campaigns (in English) that ran from August to October 2021. Interviewees voluntarily signed up via the secure web‐based survey platform REDCap hosted by Telethon Kids Institute.\\n20\\n, \\n21\\n Here, they provided demographic and contact information. They were selected if they met the CALD definition and then contacted up to three times each via phone or email to organize the interview. The CALD definition was based on the Australian Bureau of Statistics' definition,\\n22\\n and included individuals born overseas, having one or both parents born overseas from non‐English speaking countries, or whose first language is a Language Other Than English (LOTE). Interviews were conducted in English by SC over the phone or via teleconferencing software—interviewees who signed up all had sufficient English proficiency as the self‐sign‐up process was only in English.\\nDue to the high levels of English proficiency required for interview sign‐ups, we approached Ishar to partner on this research so that we could engage with CALD individuals who had lower English proficiency. Accordingly, focus group participants were recruited separately through their CALD community centre for women based in Stirling, the LGA with the highest proportion of individuals born in nonmain English‐speaking countries.\\n19\\n Women attending regular English classes were invited face‐to‐face by researchers and their teachers to participate in a focus group in lieu of their class. Focus groups were run face‐to‐face by S. J. C., G. E. and K. A. in the CALD community centre; they were conducted in English with the use of peer translators for participants with low English proficiency.\\nData were collected between August and October 2021. Of the 23 CALD individuals who voluntarily signed up online to participate in the individual interviews, we interviewed 11, with the remaining 12 being lost to follow‐up. We ran three focus groups at the Ishar multicultural centre with their female clients, comprised of 6–16 participants per group (total N = 37). All women who were present consented and participated in a focus group—there were no refusals. Interviews and focus groups lasted approximately 60 min. Of the total 48 participants, 90% were female and 29% were aged 30–39 years. Approximately, one third (35%) had been to university, 54% self‐reported high English proficiency and 27% reported having received at least their first COVID‐19 vaccine (Table 1). Further, 46% had migrated to Australia since 2010, 35% were born in Southern Asia, and 75% identified with a religion (most common being Islam and Christianity). Two of the 11 interviewees self‐identified as community leaders. Interviewees gave voluntary, informed, written consent; focus group participants gave voluntary, informed, verbal consent. All have been assigned pseudonyms.\\nSociodemographic characteristics of the 48 CALD participants, Perth, WA\\nAbbreviations: CALD, Culturally and Linguistically Diverse; TAFE, Technical and Further Education; WA, Western Australia.\\nSome participants spoke ≥1 language.\\nOther languages include Acholi, Azeri, Bomwali, Hazaragi, Hindi, Ilocano, Indonesian, Italian, Kikuyu (Bantu), Korean, Macedonian, Malay, Oromo, Somali, Tedim, Urdu and Vietnamese.\",\n \"Interview guides were developed by the multidisciplinary (project) team comprised of vaccination, medical, policy, social science and government scholars, and an indicative guide is published open‐access with our protocol.\\n12\\n The questions were simplified for the focus groups.\",\n \"All interviews and focus groups were audio‐recorded and professionally transcribed verbatim. Thematic analysis followed the Braun and Clarke method,\\n23\\n with deductive analysis in NVivo 12 using the Capability, Opportunity, Motivation, Behaviour (COM‐B) model.\\n24\\n We selected COM‐B as it is used extensively by the World Health Organization to understand reasons for (non)vaccination among specific populations.\\n25\\n G. E. generated initial codes, S. J. C. and B. N. each coded a manuscript to provide additional perspectives, including a CALD lived experience lens, and the three coders revised the matrix via a collaborative meeting before GC applied it to the full data set.\",\n \"Drawing on data from our interviews and focus groups, we created 10 subthemes encompassed within the three superordinate themes of Capability, Opportunity and Motivation (Figure 1).\\nThe Capability, Opportunity and Motivation influencing COVID‐19 vaccine uptake among Culturally and Linguistically Diverse individuals in Perth, WA. Figure modified from COM‐B model\\n24\\n; arrows indicate influence of factor on other factors.\\n[SUBTITLE] Capability [SUBSECTION] [SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n[SUBTITLE] Opportunity [SUBSECTION] [SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Motivation [SUBSECTION] Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\nMost participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\",\n \"[SUBTITLE] Knowledge about COVID‐19 and COVID‐19 vaccination [SUBSECTION] Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nMost participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n[SUBTITLE] Knowledge about how and where to get vaccinated [SUBSECTION] Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\nThough many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\",\n \"Most participants lacked an understanding of the potential severity of COVID‐19 infection, transmission routes, and vaccine ingredients or efficacy. This was particularly expressed by female focus group participants who had low English proficiency and education levels. Zi (Female, age unknown) said:I want to know, if you have your vaccines, how you know that you're protected from the virus?\\n\\nI want to know, if you have your vaccines, how you know that you're protected from the virus?\\nSome were also unsure whether the COVID‐19 vaccine brands being used in Australia would be accepted by their home country. Further, many with pre‐existing health conditions feared severe side effects from the vaccines:She has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\nShe has her two [adult] boys back in Sudan. She applied for them to come here 7–8 years ago, but they [the Australian government] didn't accept them in … and now she has so many health problem … she's scared if she take [the vaccine] … maybe something will happen to her [like death] and she will not see her [adult] kids … she doesn't know what is right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\",\n \"Though many participants knew they could receive a vaccine at a mass vaccination clinic, few were aware they could be vaccinated at a general practice clinic. Many also did not know vaccination was free. Finally, some did not know how to navigate the online booking system; a few had tried, but with no success:Yes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\\n\\nYes, I try yesterday to do online … doesn't work with me—(Ezma [Female, 40 years]; Focus Group)\",\n \"[SUBTITLE] Relationships with general practitioners (GPs) [SUBSECTION] Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\nMany participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n[SUBTITLE] Access to information/misinformation [SUBSECTION] Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\nParticipants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\\n[SUBTITLE] Access to vaccine clinics [SUBSECTION] Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\nLocations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\\n[SUBTITLE] Community influence [SUBSECTION] Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Many participants had GPs who spoke their language and/or were CALD, leading to strong trust in their recommendations:If my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\n\\nIf my doctor said [COVID‐19 vaccination] was good, I would get it—(Hana [Female, age unknown]; Focus Group)\\nAlthough many participants were unvaccinated, the appeal of convenient, trustworthy, and familiar healthcare providers meant they would choose to be vaccinated at their general practice clinics rather than travelling to mass clinics. Participants also felt more comfortable asking their CALD GP questions about the vaccine, compared to unfamiliar providers who don't speak their language. They also trusted that their GPs “don't want to kill us”. However, not all participants had established comfortable relationships with GPs, and some were concerned that GPs lack the cultural awareness skills or resources to communicate with participants.…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n…some of our [CALD] community members have got their own GPs that they trust. Some [general practice clinics are] just like going to a factory—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Participants cited significant concerns about COVID‐19 vaccine safety, often informed by misinformation. They depicted their immersion within tight‐knit communities, sharing information at gatherings, informal community events and social interactions. Friends and contacts in countries of origin or in other destination countries also spread (mis)information, including through social media.…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she doesn't have any ideas about the vaccine. And she hearing so many, you know, gossip around … she doesn't know what is the right—(Sita; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n‘Gossip’ was centred on safety: people were discussing contracting COVID‐19 from the vaccine, or extreme vaccine side effects including infertility, memory, cognition issues and death occurring specifically 2 years after vaccine receipt:…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\n\\n…almost everyone, they talking about, like, you know, the injection for the COVID, if they take it or not. And everyone said, ‘Don't take it! You're going to die! You're going to die!’—(Ezma [Female, 40 years]; Focus Group)\\nCALD individuals wanted to protect themselves against COVID‐19, but not always through vaccination, given these safety fears. Alternatives to vaccination—seen as being risk‐free—were actively shared among community members:My mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\n\\nMy mum's side [of the family] started sending … WhatsApp messages … ‘Oh, if you take an onion and you cut it in half … leave it in a corner of your house, all the coronavirus is going to go to the onion and you'll be safe’—(Inda [Female, 23 years]; Interview)\\nSome participants relied on WA government online resources and mainstream media (predominantly broadcast in English only) such as Channels Seven, Nine and the Australian Broadcasting Corporation (ABC). Few were aware of or consumed media in their language distributed by the Special Broadcasting Service (SBS), the main multilingual broadcaster in Australia. Most participants with no/low English proficiency relied heavily on social media (mostly Facebook or WhatsApp) and sourced information in their preferred language, shared by internationally residing relatives or their local CALD community.Even though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\n\\nEven though my mum can read Chinese … she wouldn't be actively going to websites to check, and she would be relying on her daughters or friends to tell her what's happening—(Liling [Female, 50 years]; Interview)\\nSome participants spoke multiple languages but still expressed a preference to consume vaccination information in only one of them—that being their first language.\\nSome participants sought accurate information from their doctors; however, their vaccine‐related fears were not always allayed.…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\n\\n…she went to the doctor but … the doctor just want to give her the vaccine. But for herself, she's not sure, you know, she confused. She want the doctor to check everything before she takes the vaccine—(Sita [Female, 29 years]; focus group attendee providing real‐time peer translation for Nell [Female, 70 years old])\\nSeveral multicultural organizations and community leaders seeking to provide accurate information attempted to collaborate with government agencies or coordinate seminars engaging the community in their own environments.…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\n\\n…she [CALD public health academic] came to the centre and gave talks about the vaccination … that's a good way of doing it—(Susi [Female, 50 years]; Interview)\\nHowever, some community leaders were themselves sharing COVID‐19 vaccine misinformation on social media:On Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nOn Facebook … even the priests and bishops are actually announcing to their parishioners the [the] vaccine is not good. (Reyna [Female, 53 years]; Community Leader, Interview)\\nParticipants were disappointed in government attempts to access and work with their communities. There were no facilitation programmes in CALD communities, who lacked resources and noted little lingual or visual diversity in promotional campaigns.[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\n\\n[Create] ‘Roll Up Your Sleeve’ [resources] in Arabic or ‘Roll Up Your Sleeve’ [resources] in Chinese. Not just English—(Kasih [Female, 50 years]; Interview)\\nWhy not approach [CALD communities] instead of asking them to come in? Because it means we give them … practical plan … the [CALD] community people … [the WA government] cannot wait for the ball, [they] need to also try to get the ball—(Audrey [Female, 43 years]; Interview)\\nParticipants believed that governments needed to be more active in reaching out to them, and said they would welcome efforts made by government and health representatives to come to the community to talk about COVID‐19 vaccination.\",\n \"Locations of mass COVID‐19 vaccination clinics were inconvenient for participants, given a lack of transport options, particularly as some could not drive. In an interview, Susi said:Even [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\n\\nEven [the mass vaccination centre] in the city [Perth convention centre] … [it's] too hard, how will people go!?\\nOther participants feared leaving their perceived safe and COVID‐19‐free community to travel into an unfamiliar part of Perth.My wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\n\\nMy wife she was really, really panicked [to leave the house] … she was kind of anxious. Same for my mum—(Tai [Male, 29 years]; Interview)\\nFor participants who did get vaccinated, the most convenient locations were clinics at or close to their workplaces or educational institutions. Some did attend mass vaccination clinics and found walk‐in processes highly convenient and well‐facilitated.\",\n \"Participants had high levels of trust in their own communities; opinions and experiences shared by peers were greatly valued and reassuring.In the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\n\\nIn the community, we were very, very scared. But after that, when I saw friend and family [had been vaccinated], a couple of people they do fine, they are perfectly fine. My friend, she's a doctor. She done vaccine last month … her husband and other colleague did it, but they got a fever … they took Panadol after … is fine—(Momo [Female, 37 years]; Focus Group)\\nHowever, the sharing of misinformation and negative opinions created situations in which some participants were torn between what their community was saying and the professional medical advice:When I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\n\\nWhen I saw, yesterday, the doctor, everybody comes to me … say don't do this … But in my heart, I say, ‘No, I'm going to take it!’—(Ezma [Female, 40 years]; Focus Group)\\nSome participants felt shame around being vaccinated, and sought to avoid undue attention or stigma from hesitant friends/family who were concerned that vaccinated people would die in 2 years:People … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\n\\nPeople … said ‘be nice with them, they going to live only two years … After two years they will die’. (Momo [Female, 37 years]; Focus Group)\\nOf the two community leaders we interviewed, one was completely against the idea of any medical intervention, including vaccination:I've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\n\\nI've never taken any flu vaccines, I never take any painkillers, I avoid doctors like the plague … I want my body to learn how to cope and how to fight those diseases … And if I can't cope, I will simply go to the next world—(Ubel [Male, 78 years]; Interview)\\nUbel volunteers at a community organization for refugees and migrants in Perth. His role is to share information with those in the database. When asked whether he shares information on COVID‐19 vaccination, he said:We get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\n\\nWe get a quite a number [of emails from various official and unofficial sources] and I've passed on. Except yesterday for the first time, I did not send one on because this [particular] group sent me an email and they are totally convinced about COVID and it was a bit too strong for my liking. So I did not send it, but that's the first one.\\nSome participants also discussed the role of community leaders, saying these leaders did not promote that they were vaccinated, fearing negative feedback or loss of power:[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\n[If community leaders] had a vaccine, they don't tell people, their own group … because other people … look up to them and … most of us don't believe in [the vaccine]—(Reyna [Female, 53 years]; Community Leader, Interview)\\nReyna was a pro‐vaccination community leader, and she expressed frustration in feeling solely responsible for vaccine uptake in her wider CALD community. She felt the government had not come to her organization to provide support; community leaders instead had to organize seminars themselves with health representatives to answer community questions.We're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\\n\\nWe're not the only not‐for‐profit community centre that are not well informed of the COVID and now [the WA government is] asking us to ‘roll up’ … how could you have a vaccine if you weren't knowledgeable, you have no education about it at all?—(Reyna [Female, 53 years]; Community Leader, Interview)\",\n \"Most participants were “scared”, “confused” or “unsure” about COVID‐19 vaccines, particularly because, as Umme said in a focus group “this medicine only come in 2 years, it come quickly, why so quickly, maybe something wrong”. Yet despite these concerns, many were nevertheless willing to be vaccinated to protect themselves, their family and their community.I'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\n\\nI'm getting this so that others around me don't get sick. I'm getting this so that I can protect my mum, I can protect my dad, I can protect my friends—(Angelo [Male, 37 years]; Interview)\\nParticipants, aware that “Corona is coming”, felt they would be at greater risk once international and interstate borders opened, declaring that they would get vaccinated once WA borders opened (but likely not before). Others said the border opening would mean they could visit their friends and family overseas; they believed COVID‐19 vaccination would be a requirement for this:I want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\\n\\nI want to go visit my family [in Ethiopia] … But [also] I'm happy that we are closed here. Because then we … stay safe—(Mahsa [Female, 34 years]; Focus Group)\",\n \"This is the first study to explore perceptions and attitudes to the COVID‐19 vaccine, barriers to preventing and strategies promoting vaccine uptake among CALD populations in Perth, WA. Similar to studies from the United Kingdom\\n11\\n, \\n26\\n and New South Wales,\\n8\\n we found high levels of COVID‐19 vaccine hesitancy among CALD individuals due to safety concerns, lack of understanding and high levels of misinformation perpetuated by CALD peers or social media platforms. Hesitancy and low uptake were particularly apparent among participants with low levels of education and low English proficiency. These findings have since been confirmed in a large regression analysis of over 12 million Australians, which identified that those with low levels of education and English proficiency were less likely to have received their first COVID‐19 vaccine dose.\\n27\\n\\n\\nDespite their concerns, participants were not completely opposed to vaccination, and we were struck by how they did not see the COVID‐19 vaccination programme as being for people like them. They were exposed to misinformation through their social networks, as previous studies have also found,\\n11\\n and the hesitancy this generated is a wider problem. However, participants also felt that governments were not forthcoming with resources to address their concerns. Participants were ostensibly included in cohorts that were eligible for vaccination at the time of data collection, and some of them had been eligible many months prior, due to their age. However, although they were ostensibly being asked to ‘roll up’, they received few or no personal cues to avail themselves of vaccinations.\\nPractical barriers were also evident. As previously identified by Abdi et al.,\\n28\\n refugees and migrants tend to have quite demanding competing priorities, such as seeking employment and accommodation and thus need prompting for vaccination. In some cases, our participants' efforts to be vaccinated were thwarted by the online booking system. In fact, some of our research team assisted participants to book vaccination appointments on their smartphones after the focus groups or provide further information to partners who waited in the car park to speak with them. It is not just our participants who struggled with the online booking system: Seale et al.\\n29\\n describe a ‘digital divide’ in which those without access to the internet and have low English proficiency are disadvantaged in their ability to access COVID‐19 vaccines, and Biddle et al.\\n27\\n identified that Australians who completed the 2016 census on paper rather than online had a lower probability of vaccination, suggesting that individuals with lower digital literacy may experience more barriers to vaccination. Booking systems must be accessible to varying levels of digital literacy and English proficiency.\\nSocial norms are a strong influence on vaccine uptake.\\n16\\n, \\n30\\n For CALD people in Perth, social norms were blocking COVID‐19 vaccination. Vaccinated CALD individuals were not sharing their experiences or reasons for vaccinating, facilitating ongoing shame and stigma, and thereby perpetuating continued silence for those who did vaccinate; tendencies that have also been found in local ‘alternative’ communities with regard to childhood vaccination.\\n31\\n Some CALD leaders were also not actively encouraging vaccination due to their own concerns or opposition, an issue identified early on by the Australian multicultural sector as a potential barrier.\\n32\\n\\n\\nAnother well‐established influence on vaccine uptake is receiving a recommendation from a healthcare provider.\\n30\\n Our participants showed great trust in their GPs and reported positive experiences of consultation. Accordingly, they believed that GPs would have more opportunities to address COVID‐19 vaccine concerns than mass vaccination clinic staff. However, GPs may not always be easily accessible within an acceptable timeframe to the patient,\\n33\\n meaning some CALD individuals are turning to or relying on social media for their health/COVID‐19 information.\\n9\\n\\n\\nOur participants' high levels of trust in governing authorities was a novel finding, mostly related to feeling safe from COVID‐19 in WA. Focus groups with CALD individuals in New South Wales instead found participants were concerned that the government was using COVID‐19 vaccination to control people.\\n8\\n We were struck by how our participants' trust in government emanated from people whose experiences in their countries of origin had often been traumatic, culminating in their refugee status.\",\n \"A strong motivation for this research was to assist in strategic design, informing ongoing and future implementation of public health initiatives among CALD populations. Our experiences engaging with our CALD participants and their partners, including helping them to book appointments after the focus groups, emphasized how important it is that governments promptly address vaccine access issues, as well as campaigning and working directly with CALD groups, and featuring them visibly in messaging. The barriers regarding lack of access to resources and the circulation of misinformation must be addressed as soon as possible to reduce the COVID‐19 burden amongst the CALD population in WA. These lessons can then be utilized for future vaccine doses, for reaching the children of CALD populations, and for wider public health initiatives.\\nAt the time of our study, CALD participants and their communities were not publicly sharing their vaccinated status and proudly promoting COVID‐19 vaccinations. However, it is unreasonable to demand more social courage from CALD individuals when what they need is more support. Given that governments are responsible for making vaccination accessible and acceptable to improve and sustain uptake,\\n34\\n they must recognize, work with, and train influential community leaders on the importance of and how to have meaningful vaccine conversations with their peers.\\n32\\n A recent review identified the who, the what and the how for effective community engagement during a pandemic: essentially, it is important to engage with the community from the very beginning for a better chance of behaviour change within the community.\\n35\\n In eastern Australia, Mahimbo et al.\\n9\\n identified that CALD communities are eager for community engagement, and for their leaders to champion the vaccine. Seale et al.\\n29\\n identified that while some CALD organizations in Australia were updating local CALD leaders on relevant COVID‐19 information to share with their community, there was a lack of training in how to do so. Training influential community leaders can occur over a relatively short period of time: a COVID‐19 information and support programme was offered in the United States to CALD community leaders which involved 1‐hour conference calls twice a week (we acknowledge that this is then reliant on the leaders having a decent level of computer literacy.)\\n29\\n Further, given the trust and respect our participants held for their health practitioners, it is imperative that governments support GPs to effectively communicate about COVID‐19 vaccination with CALD patients; this requires resources, training and remuneration.\\n36\\n Since our data collection, the Australian National Centre for Immunisation Research and Surveillance have published a COVID‐19 decision aid in Greek, Vietnamese, Arabic, Simplified Chinese and Traditional Chinese.\\n37\\n Immunisation providers could use or share this tool with their CALD patients (though its effectiveness in increasing vaccine uptake is yet to be evaluated).\\nOur participants' high levels of trust in governing authorities is an untapped opportunity. WA participants knew that ‘Corona is coming’. Now that it has arrived, the government and healthcare providers must maintain the trust of CALD by reaching them promptly with vaccine doses and culturally appropriate information to reduce the burden of COVID‐19. If these measures fail, it would be entirely reasonable for CALD individuals to not blame their own choices, and instead look to the authorities they believed would protect them and ask why they were not offered greater support and access.\\n[SUBTITLE] Limitations [SUBSECTION] Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\\nQualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\",\n \"Qualitative data is not widely generalizable, as its function is to generate deep insights when little is known about an issue, or when data cannot be captured appropriately via quantitative means, including across language barriers. Different Australian states had different pandemic experiences and our data may not be applicable to other states or countries, although our insights and reflections remain pertinent. Our interviews and focus groups took place well into the COVID‐19 pandemic and vaccine rollout, but before community transmission in WA—the subsequent ‘first wave’ from February 2022 may have changed participants' views. Our recruitment method for interviews skewed demographics towards those with high English proficiency, females (despite a social media campaign specifically recruiting men), high levels of education and high COVID‐19 vaccine uptake. However, we mitigated this by collaborating with a community centre for recent migrants/refugees to undertake focus groups with women possessing low levels of English proficiency, education, and COVID‐19 vaccine uptake. Participants' language constraints were mitigated by strong peer relationships that facilitated peer translation. Despite our efforts to mitigate these limitations, our study does not represent the entire CALD community in Perth.\",\n \"This study provides in‐depth insight into the perceptions and attitudes of CALD communities in Perth, WA, regarding the COVID‐19 vaccine, identifying facilitators and barriers for engagement and uptake. CALD participants faced barriers including a lack of knowledge about COVID‐19 and the vaccines, low English proficiency and education levels, misinformation, passive government communication strategies and limited access to vaccine clinics/providers. They were, however, motivated to vaccinate by the imminent opening of state and international borders, trust in government and healthcare authorities, travel intentions and the desire to protect themselves and others. Despite high levels of trust and significant desire for vaccines, strategies at the time of our study were not meeting the CALD community's needs, and thus those within it remain at risk from the disease. Tailored intervention strategies are required to provide knowledge, address misinformation, and facilitate access to ensure uptake of COVID‐19 vaccines—including for additional doses—amongst CALD communities. Governments must work with trusted CALD community members and immunization providers such as GPs to disseminate tailored COVID‐19 vaccine information and adequately translated resources so that community members themselves can become stronger vaccine ambassadors to each other.\",\n \"Samantha J. Carlson supervised the data collection and analysis, led the initial manuscript drafting, contributed to the study design, contributed to project administration and coordinated the ethics and stakeholder engagement. Gracie Edwards collected the data and conducted the analysis as well as drafted the literature review and contributed to the overall manuscript drafting. Christopher C. Blyth contributed to the conceptualization of the project and provided oversight and input into the study and funding attainment, and reviewed and edited the manuscript. Barbara Nattabi supervised the data collection and analysis, contributed to the coding framework, provided CALD expertise, contributed to project administration and reviewed and edited the manuscript. Katie Attwell conceptualized the project, co‐developed the broader study methodology with Christopher C. Blyth, led the funding attainment, and supervised the conduct and administration of the broader research project. She contributed to the literature review and the drafting, reviewing and editing of the manuscript. All authors have approved the final manuscript.\",\n \"K. A. and C. C. B. are specialist advisors to the Australian Technical Advisory Group on Immunisation. The remaining authors declare no conflict of interest.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,"discussion",null,null,"conclusions",null,"COI-statement"],"string":"[\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n \"conclusions\",\n null,\n \"COI-statement\"\n]"},"keywords":{"kind":"list like","value":["attitudes","COVID‐19","ethnic and racial minorities","health knowledge","immunization","misinformation","vaccination"],"string":"[\n \"attitudes\",\n \"COVID‐19\",\n \"ethnic and racial minorities\",\n \"health knowledge\",\n \"immunization\",\n \"misinformation\",\n \"vaccination\"\n]"}}},{"rowIdx":357,"cells":{"title":{"kind":"string","value":"Impact of COVID-19 on HIV late diagnosis in a specialized German centre."},"pmid":{"kind":"string","value":"36263724"},"background_abstract":{"kind":"string","value":"The ongoing COVID-19 pandemic has been impeding HIV diagnosis and treatment worldwide. Data on the impact of COVID-19 on late diagnosis (LD) in Germany are lacking. Here we present novel data of a single-centre German HIV cohort assessing LD during COVID-19."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"This is a non-interventional, single-centre retrospective cohort assessing the rate of LD comparing HIV diagnoses pre-COVID-19 with those during the COVID-19 pandemic. New diagnoses between 1 January 2019 and 1 February 2020 were classified as pre-COVID-19, and diagnoses between 1 February 2020 and 1 October 2021 were classified as during COVID-19."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Between 1 January 2019 and 1 October 2021, 75 patients presented with newly diagnosed HIV infection, 34 pre-COVID-19 and 41 during COVID-19. LD increased to 83% (n = 34/41) during COVID-19 versus 59% (n = 20/34) pre-COVID-19, and CDC stage C3 rose to 44% (n = 18) versus 27%. Hospitalization rate increased to 49% (n = 20) during COVID-19 versus 29% pre-COVID-19, and 12% (n = 5) presented with HIV-associated neurological disease, whereas none were observed in the pre-COVID-19 group. The incidence of LD (p = 0.020), CD4 count < 350 cells/μL (p = 0.037) and < 200 cells/μL (p = 0.022) were statistically significantly associated with the ongoing COVID-pandemic. An association with HIV transmission risk was borderline significant (p = 0.055)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Despite comparable annual rates of new HIV diagnoses, LD has been increasing during the COVID-19 pandemic, resulting in more opportunistic infections and higher hospitalization rates, possibly reflecting pandemic-related shortages in HIV testing and care facilities. Maintaining HIV testing opportunities and access to treatment during a pandemic is crucial so as not to impede WHO elimination goals and so as to prevent an increase in AIDS-related morbidity and mortality."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Delayed Diagnosis","HIV Infections","COVID-19","Retrospective Studies","Pandemics"],"string":"[\n \"Humans\",\n \"Delayed Diagnosis\",\n \"HIV Infections\",\n \"COVID-19\",\n \"Retrospective Studies\",\n \"Pandemics\"\n]"},"pmcid":{"kind":"string","value":"9874758"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"The study was designed as a non‐interventional single‐centre retrospective cohort at the University Hospital Bonn, a maximum healthcare provider with a specialized infectious disease outpatient clinic as well as an infectious disease ward. The aim was to assess the rate of LD, comparing HIV diagnoses pre‐COVID‐19 (PC) with those during the COVID‐19 pandemic (DC). LD was defined as a first CD4 count < 350 cells/μL with or without an AIDS‐defining illness at the time of HIV diagnosis. New diagnoses between 1 January 2019 and 1 February 2020 were classified as PC patients, and diagnoses between 1 February 2020 and 1 October 2021 were classified as DC patients. In addition to assessment of demographic data, mode of HIV transmission, CD4 count, HIV‐RNA and concomitant opportunistic infections (OIs), the hospitalization rate and possible necessary intensive care treatment and mortality were assessed. CD4 count was assessed by flow cytometry using XN‐Series (Sysmex America Inc., Lincolnshire, IL, USA) at the Institute of Clinical Biochemistry, University Hospital Bonn. Through use of the reagent Fluorocell WDF (Sysmex Europe GmbH, Norderstedt, Germany), absolute count (number/μL) and relative count (%) of CD4 and CD8 were assessed. HIV‐RNA was measured with an Abbot m2000 assay, with a lower detection rate of < 40 copies/mL blood.\n[SUBTITLE] Statistical Analysis [SUBSECTION] Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki.\nStatistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"Baseline characteristics are shown in Table 1. Between 1 January 2019 and 1 October 2021, 75 patients presented with newly diagnosed HIV infection at the University Hospital Bonn, 34 PC and 41 DC. Median ages were 33 years (range 18–71) for PC presentation and 34 years (18–76) for DC presentation. The main route of transmission was men who have sex with men (MSM), which decreased from 71% (n = 24/34) for PC patients to 51% (n = 21/41) for DC patients. The rate of first diagnosis during hospital stay was found to be highest among DC patients (46%, n = 19/41) and was greater than among PC patients, 30% (n = 10/33) of whom were diagnosed during hospital admission. Diagnosis through testing at specialist care (e.g. haematologist, dermatologist, gynaecologist, infectious disease specialist) due to referral from a general practitioner was the most common diagnostic method among PC patients (39%, n = 13/33).\nBaseline characteristics\nAbbreviation: MSM, men who have sex with men.\nLate presentation increased to 83% (n = 34/41) for DC patients versus 59% (n = 20/34) for PC patients, and CDC stage C3 rose to 44% (n = 18/41) versus 27% (n = 9/34), respectively. Median CD4 count was 291 cells/μL (range 3–930) for PC patients and decreased to 118 cells/μL (10–782) for DC patients. Consistently, median HIV‐RNA almost doubled in DC patients, with a median of 93.759 copies/mL (range 408–9110.000) versus 48.598 copies/mL (range 957–8012.596) for PC patients; 46% (n = 19/41) of DC patients presented with AIDS‐defining illness versus 26% (n = 9/34) of PC patients. Pneumocystis jirovecii pneumonia (PJP) was the most often reported AIDS‐defining illness and was found more frequently among PC (67%, n = 6/9) than among DC patients (37%, n = 7/19). The rate of hospitalization increased to 49% (n = 20/41) in DC patients versus 29% (n = 10/34) in PC patients. Thus, the proportion of patients who required intensive care treatment was higher among the PC cohort (40%, n = 4/10) than among the DC cohort (30%, n = 6/20). Interestingly, 26% (n = 5/19) of DC patients with AIDS‐defining illness presented with AIDS‐associated neurological disease, including cerebral toxoplasmosis and HIV‐associated neurological disorder, whereas none were observed among PC patients. In both groups the death rate was low (3%, n = 1/34 for PC and 2%, n = 1/41 for DC); one patient died due to PJP and one died due to lymphoma.\nThe incidence of LD (p = 0.020), CD4 count < 350 cells/μL (p = 0.037) and < 200 cells /μL (p = 0.022) were statistically significantly associated with the ongoing COVID pandemic. Association with transmission risk was borderline significant (p = 0.055)."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Statistical Analysis","AUTHOR CONTRIBUTIONS"],"string":"[\n \"INTRODUCTION\",\n \"Statistical Analysis\",\n \"AUTHOR CONTRIBUTIONS\"\n]"},"other_sections_texts":{"kind":"list like","value":["Late diagnosis (LD) in HIV remains a challenging problem despite widespread accessibility to diagnostic tests and combined antiretroviral therapy (cART). In Germany, about one‐third of people living with HIV are diagnosed late, as defined by CD4 T‐cell count < 350 cells/μL with/or without manifest AIDS‐defining illness [1]. LD is known to increase morbidity and mortality in affected in‐patients and, moreover, to increase healthcare costs as well as transmission risk as a result of individuals being unaware of their HIV status [2, 3]. The ongoing COVID‐19 pandemic has been impeding HIV diagnosis and treatment worldwide [4].\nAs there has been substantial progress in decreasing the number of new HIV infections through low‐threshold HIV testing, treatment as prevention (TasP) and initiation of HIV pre‐exposure prophylaxis, the ongoing COVID‐19 pandemic has raised major concerns in keeping the high‐quality standard accessible, which is potentially increasing the rate of HIV transmissions and additionally increasing the number of HIV late diagosis.\nCurrently, data on the impact of COVID‐19 on late diagnosis in Germany are lacking. Here we present novel data of a single‐centre German HIV cohort assessing the rate and impact of late diagnosis during COVID‐19.","Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki.","KvB designed the study and concept. Sample collection was performed by all co‐authors. Data analysis was performed by KvB. The manuscript was written by KvB and CB with full input from all co‐authors."],"string":"[\n \"Late diagnosis (LD) in HIV remains a challenging problem despite widespread accessibility to diagnostic tests and combined antiretroviral therapy (cART). In Germany, about one‐third of people living with HIV are diagnosed late, as defined by CD4 T‐cell count < 350 cells/μL with/or without manifest AIDS‐defining illness [1]. LD is known to increase morbidity and mortality in affected in‐patients and, moreover, to increase healthcare costs as well as transmission risk as a result of individuals being unaware of their HIV status [2, 3]. The ongoing COVID‐19 pandemic has been impeding HIV diagnosis and treatment worldwide [4].\\nAs there has been substantial progress in decreasing the number of new HIV infections through low‐threshold HIV testing, treatment as prevention (TasP) and initiation of HIV pre‐exposure prophylaxis, the ongoing COVID‐19 pandemic has raised major concerns in keeping the high‐quality standard accessible, which is potentially increasing the rate of HIV transmissions and additionally increasing the number of HIV late diagosis.\\nCurrently, data on the impact of COVID‐19 on late diagnosis in Germany are lacking. Here we present novel data of a single‐centre German HIV cohort assessing the rate and impact of late diagnosis during COVID‐19.\",\n \"Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\\nThe study was conducted according to the Declaration of Helsinki.\",\n \"KvB designed the study and concept. Sample collection was performed by all co‐authors. Data analysis was performed by KvB. The manuscript was written by KvB and CB with full input from all co‐authors.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null],"string":"[\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Statistical Analysis","RESULTS","DISCUSSION","AUTHOR CONTRIBUTIONS"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Statistical Analysis\",\n \"RESULTS\",\n \"DISCUSSION\",\n \"AUTHOR CONTRIBUTIONS\"\n]"},"all_sections_texts":{"kind":"list like","value":["Late diagnosis (LD) in HIV remains a challenging problem despite widespread accessibility to diagnostic tests and combined antiretroviral therapy (cART). In Germany, about one‐third of people living with HIV are diagnosed late, as defined by CD4 T‐cell count < 350 cells/μL with/or without manifest AIDS‐defining illness [1]. LD is known to increase morbidity and mortality in affected in‐patients and, moreover, to increase healthcare costs as well as transmission risk as a result of individuals being unaware of their HIV status [2, 3]. The ongoing COVID‐19 pandemic has been impeding HIV diagnosis and treatment worldwide [4].\nAs there has been substantial progress in decreasing the number of new HIV infections through low‐threshold HIV testing, treatment as prevention (TasP) and initiation of HIV pre‐exposure prophylaxis, the ongoing COVID‐19 pandemic has raised major concerns in keeping the high‐quality standard accessible, which is potentially increasing the rate of HIV transmissions and additionally increasing the number of HIV late diagosis.\nCurrently, data on the impact of COVID‐19 on late diagnosis in Germany are lacking. Here we present novel data of a single‐centre German HIV cohort assessing the rate and impact of late diagnosis during COVID‐19.","The study was designed as a non‐interventional single‐centre retrospective cohort at the University Hospital Bonn, a maximum healthcare provider with a specialized infectious disease outpatient clinic as well as an infectious disease ward. The aim was to assess the rate of LD, comparing HIV diagnoses pre‐COVID‐19 (PC) with those during the COVID‐19 pandemic (DC). LD was defined as a first CD4 count < 350 cells/μL with or without an AIDS‐defining illness at the time of HIV diagnosis. New diagnoses between 1 January 2019 and 1 February 2020 were classified as PC patients, and diagnoses between 1 February 2020 and 1 October 2021 were classified as DC patients. In addition to assessment of demographic data, mode of HIV transmission, CD4 count, HIV‐RNA and concomitant opportunistic infections (OIs), the hospitalization rate and possible necessary intensive care treatment and mortality were assessed. CD4 count was assessed by flow cytometry using XN‐Series (Sysmex America Inc., Lincolnshire, IL, USA) at the Institute of Clinical Biochemistry, University Hospital Bonn. Through use of the reagent Fluorocell WDF (Sysmex Europe GmbH, Norderstedt, Germany), absolute count (number/μL) and relative count (%) of CD4 and CD8 were assessed. HIV‐RNA was measured with an Abbot m2000 assay, with a lower detection rate of < 40 copies/mL blood.\n[SUBTITLE] Statistical Analysis [SUBSECTION] Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki.\nStatistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki.","Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\nThe study was conducted according to the Declaration of Helsinki.","Baseline characteristics are shown in Table 1. Between 1 January 2019 and 1 October 2021, 75 patients presented with newly diagnosed HIV infection at the University Hospital Bonn, 34 PC and 41 DC. Median ages were 33 years (range 18–71) for PC presentation and 34 years (18–76) for DC presentation. The main route of transmission was men who have sex with men (MSM), which decreased from 71% (n = 24/34) for PC patients to 51% (n = 21/41) for DC patients. The rate of first diagnosis during hospital stay was found to be highest among DC patients (46%, n = 19/41) and was greater than among PC patients, 30% (n = 10/33) of whom were diagnosed during hospital admission. Diagnosis through testing at specialist care (e.g. haematologist, dermatologist, gynaecologist, infectious disease specialist) due to referral from a general practitioner was the most common diagnostic method among PC patients (39%, n = 13/33).\nBaseline characteristics\nAbbreviation: MSM, men who have sex with men.\nLate presentation increased to 83% (n = 34/41) for DC patients versus 59% (n = 20/34) for PC patients, and CDC stage C3 rose to 44% (n = 18/41) versus 27% (n = 9/34), respectively. Median CD4 count was 291 cells/μL (range 3–930) for PC patients and decreased to 118 cells/μL (10–782) for DC patients. Consistently, median HIV‐RNA almost doubled in DC patients, with a median of 93.759 copies/mL (range 408–9110.000) versus 48.598 copies/mL (range 957–8012.596) for PC patients; 46% (n = 19/41) of DC patients presented with AIDS‐defining illness versus 26% (n = 9/34) of PC patients. Pneumocystis jirovecii pneumonia (PJP) was the most often reported AIDS‐defining illness and was found more frequently among PC (67%, n = 6/9) than among DC patients (37%, n = 7/19). The rate of hospitalization increased to 49% (n = 20/41) in DC patients versus 29% (n = 10/34) in PC patients. Thus, the proportion of patients who required intensive care treatment was higher among the PC cohort (40%, n = 4/10) than among the DC cohort (30%, n = 6/20). Interestingly, 26% (n = 5/19) of DC patients with AIDS‐defining illness presented with AIDS‐associated neurological disease, including cerebral toxoplasmosis and HIV‐associated neurological disorder, whereas none were observed among PC patients. In both groups the death rate was low (3%, n = 1/34 for PC and 2%, n = 1/41 for DC); one patient died due to PJP and one died due to lymphoma.\nThe incidence of LD (p = 0.020), CD4 count < 350 cells/μL (p = 0.037) and < 200 cells /μL (p = 0.022) were statistically significantly associated with the ongoing COVID pandemic. Association with transmission risk was borderline significant (p = 0.055).","To our knowledge this is the first real‐life cohort presenting novel data on the impact of COVID‐19 in HIV late diagnosis in Germany.\nEven with easy accessibility to testing facilities and the prompt treatment of newly diagnosed HIV patients, 40–55% of patients are still diagnosed with late diagnosis in Germany [5, 6]. Other European cohorts have shown similar LD rates between 42% and 58% [7, 8, 9, 10]. In the COHERE cohort, being heterosexual, originating from southern Europe or being of African descent were found to be associated with the highest risk of LD [10]. Older age has also been associated with an increased risk of LD [5]. It is an urgent goal to diagnose HIV patients at an early stage as it is a known fact that LD leads to increased morbidity and mortality as well as increased health costs. The UNAIDS 95‐95‐95 goals are at increasing risk of being missed due to the ongoing focus on COVID‐19 and the resulting missed opportunities of treating HIV patients. As reported in the UNAIDS HIV Services Tracking, a large, sustained decrease in HIV testing in the reporting countries was observed [11].\nWith the beginning of the COVID‐19 pandemic in Germany in February 2020, we found the rate of LD to be rising at our infectious disease department. The overall rate of LD during the ongoing pandemic has shown a 40% increase over the pre‐pandemic time‐frame, resulting in a 3.4‐fold higher odds ratio (95% confidence interval: 1.18–9.83). This is most likely due to the numerous impacts of the ongoing COVID‐9 pandemic in Germany. Beginning with increasing numbers of SARS‐CoV‐2 infections in spring 2020, the fear of acquiring SARS‐CoV‐2 infection rose steadily, which led to an overall decreased consultation rate by, and access to, physicians. Therefore, possible symptoms of progressed HIV infection might have been neglected or their investigation postponed by the individual, or been misinterpreted due to the main focus on COVID‐19.\nAccess to low‐threshold testing was impeded by down‐regulation of community service testing at checkpoints and municipal health centres. As a result, two main areas for detecting undiagnosed HIV infection (whether with symptoms or asymptomatic) were disrupted.\nConcerning risk of transmission, we found a decreasing rate of new HIV infections in MSM which was also observed in an Italian cohort during the ongoing COVID‐19 pandemic [12]. This could be due to reduced contacts during lockdown as well as the fewer possibilities for testing mentioned earlier, which is more widespread and implemented in this risk group. Previous cohorts have shown a decrease in late diagnosis in MSM over time, which might be due to the implementation of pre‐exposure prophylaxis [10]. In our study we found an increase in LD among heterosexuals. Moreover, transmission via heterosexual contact has been described earlier as an associated risk factor for LD [11]. Indeed, this group might be more affected by having less information about testing strategies as well as by a lack of awareness regarding their individual risk of infection and could therefore have been even more impeded during the ongoing COVID‐19 pandemic.\nRegarding the institution in which HIV infection was diagnosed, we found a shift towards diagnosis in hospital due to mandatory treatment, whereas diagnosis before COVID‐19 was most often performed in specialist practices. However, due to the high rate of LD in our cohort, this finding has to be interpreted with caution. This reflects the fact that ongoing LD during the COVID‐19 pandemic might be due to the reduced availability of low‐threshold testing and reduced slots at general practitioners. Moreover, LD, with its concomitant laboratory and clinical abnormalities, requires special awareness, underlining the fact that all physicians, regardless of their speciality, need to be aware of HIV‐associated symptoms and disease in order to determine the HIV diagnosis.\nA low CD4 count reflects the impaired immune function and increased risk of developing AIDS‐defining diseases, which all lead to increased morbidity and mortality. In our study, CD4 counts < 350 cells/μL and < 200 cells/μL were statistically significantly associated with the ongoing COVID‐19 pandemic, underlining the fact that we need to put all our efforts towards diagnosing patients at an earlier stage in order to prevent severe immune dysfunction and subsequent AIDS‐defining diseases.\nAt the same time, we found an almost doubled rate of AIDS‐defining diseases during the ongoing COVID‐19 pandemic. PJP, which is the most prevalent AIDS‐defining disease, was found most often in our study. It is important to note that more than half of patients with PJP are so severely ill that they require intensive care treatment; 40% (n = 2/5) of DC patients with PJP who required intensive care treatment had to undergo mechanical ventilation via intubation, as compared with 25% (n = 1/4) of PC patients. The overall increased rate of AIDS‐defining diseases raises morbidity and mortality, which also increases the cost of healthcare. Moreover, persistent organ damage or development of renal insufficiency after recovering from PJP or other AIDS‐defining diseases has to be taken into consideration.\nInterestingly, we diagnosed several patients with AIDS‐related neurological disorders, whereas none had been observed in the comparable pre‐COVID cohort. Since implementation of cART, incidence of HIV‐associated neurological diseases has been decreasing, and thus patients with LD are still at an elevated risk [13]. In particular, AIDS‐related neurological disorders cause significant and, most often, permanent harm to the patient and therefore early diagnosis is of vital importance. Both of our patients with cerebral toxoplasmosis still have tremendous neurological deficits. One patient who presented with hemiplegia due to cerebral toxoplasmosis and severely spread Kaposi sarcoma even went blind due to cytomegalovirus‐associated retinitis.\nOur study has several limitations. First, it is a monocentric study, and second it only reflects data from two time slots which are close together. The overall higher rate of LD might be overestimated due to coincidence and the historical fact of the increased amount of LD in our hospital, as it is well known for its special infectious disease department. Thus, this study reflects a real‐life cohort being affected by the ongoing COVID‐19 pandemic.\nIn conclusion we found an increased rate of LD in our cohort during the persisting COVID‐19 pandemic despite stable annual numbers of new HIV diagnoses, which underlines the barriers to HIV testing as well as general care as a result of the focus on COVID‐19. The outcome is more AIDS‐defining diseases and higher hospitalization rates. Maintaining HIV testing opportunities and access to treatment during a pandemic is crucial if it is not to impede WHO elimination goals and we are to prevent an increase in AIDS‐related morbidity and mortality.","KvB designed the study and concept. Sample collection was performed by all co‐authors. Data analysis was performed by KvB. The manuscript was written by KvB and CB with full input from all co‐authors."],"string":"[\n \"Late diagnosis (LD) in HIV remains a challenging problem despite widespread accessibility to diagnostic tests and combined antiretroviral therapy (cART). In Germany, about one‐third of people living with HIV are diagnosed late, as defined by CD4 T‐cell count < 350 cells/μL with/or without manifest AIDS‐defining illness [1]. LD is known to increase morbidity and mortality in affected in‐patients and, moreover, to increase healthcare costs as well as transmission risk as a result of individuals being unaware of their HIV status [2, 3]. The ongoing COVID‐19 pandemic has been impeding HIV diagnosis and treatment worldwide [4].\\nAs there has been substantial progress in decreasing the number of new HIV infections through low‐threshold HIV testing, treatment as prevention (TasP) and initiation of HIV pre‐exposure prophylaxis, the ongoing COVID‐19 pandemic has raised major concerns in keeping the high‐quality standard accessible, which is potentially increasing the rate of HIV transmissions and additionally increasing the number of HIV late diagosis.\\nCurrently, data on the impact of COVID‐19 on late diagnosis in Germany are lacking. Here we present novel data of a single‐centre German HIV cohort assessing the rate and impact of late diagnosis during COVID‐19.\",\n \"The study was designed as a non‐interventional single‐centre retrospective cohort at the University Hospital Bonn, a maximum healthcare provider with a specialized infectious disease outpatient clinic as well as an infectious disease ward. The aim was to assess the rate of LD, comparing HIV diagnoses pre‐COVID‐19 (PC) with those during the COVID‐19 pandemic (DC). LD was defined as a first CD4 count < 350 cells/μL with or without an AIDS‐defining illness at the time of HIV diagnosis. New diagnoses between 1 January 2019 and 1 February 2020 were classified as PC patients, and diagnoses between 1 February 2020 and 1 October 2021 were classified as DC patients. In addition to assessment of demographic data, mode of HIV transmission, CD4 count, HIV‐RNA and concomitant opportunistic infections (OIs), the hospitalization rate and possible necessary intensive care treatment and mortality were assessed. CD4 count was assessed by flow cytometry using XN‐Series (Sysmex America Inc., Lincolnshire, IL, USA) at the Institute of Clinical Biochemistry, University Hospital Bonn. Through use of the reagent Fluorocell WDF (Sysmex Europe GmbH, Norderstedt, Germany), absolute count (number/μL) and relative count (%) of CD4 and CD8 were assessed. HIV‐RNA was measured with an Abbot m2000 assay, with a lower detection rate of < 40 copies/mL blood.\\n[SUBTITLE] Statistical Analysis [SUBSECTION] Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\\nThe study was conducted according to the Declaration of Helsinki.\\nStatistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\\nThe study was conducted according to the Declaration of Helsinki.\",\n \"Statistical analysis was performed with IBM SPSS 28. Fisher's exact, χ2 and Mann–Whitney U‐test were used for statistical analysis.\\nThe study was conducted according to the Declaration of Helsinki.\",\n \"Baseline characteristics are shown in Table 1. Between 1 January 2019 and 1 October 2021, 75 patients presented with newly diagnosed HIV infection at the University Hospital Bonn, 34 PC and 41 DC. Median ages were 33 years (range 18–71) for PC presentation and 34 years (18–76) for DC presentation. The main route of transmission was men who have sex with men (MSM), which decreased from 71% (n = 24/34) for PC patients to 51% (n = 21/41) for DC patients. The rate of first diagnosis during hospital stay was found to be highest among DC patients (46%, n = 19/41) and was greater than among PC patients, 30% (n = 10/33) of whom were diagnosed during hospital admission. Diagnosis through testing at specialist care (e.g. haematologist, dermatologist, gynaecologist, infectious disease specialist) due to referral from a general practitioner was the most common diagnostic method among PC patients (39%, n = 13/33).\\nBaseline characteristics\\nAbbreviation: MSM, men who have sex with men.\\nLate presentation increased to 83% (n = 34/41) for DC patients versus 59% (n = 20/34) for PC patients, and CDC stage C3 rose to 44% (n = 18/41) versus 27% (n = 9/34), respectively. Median CD4 count was 291 cells/μL (range 3–930) for PC patients and decreased to 118 cells/μL (10–782) for DC patients. Consistently, median HIV‐RNA almost doubled in DC patients, with a median of 93.759 copies/mL (range 408–9110.000) versus 48.598 copies/mL (range 957–8012.596) for PC patients; 46% (n = 19/41) of DC patients presented with AIDS‐defining illness versus 26% (n = 9/34) of PC patients. Pneumocystis jirovecii pneumonia (PJP) was the most often reported AIDS‐defining illness and was found more frequently among PC (67%, n = 6/9) than among DC patients (37%, n = 7/19). The rate of hospitalization increased to 49% (n = 20/41) in DC patients versus 29% (n = 10/34) in PC patients. Thus, the proportion of patients who required intensive care treatment was higher among the PC cohort (40%, n = 4/10) than among the DC cohort (30%, n = 6/20). Interestingly, 26% (n = 5/19) of DC patients with AIDS‐defining illness presented with AIDS‐associated neurological disease, including cerebral toxoplasmosis and HIV‐associated neurological disorder, whereas none were observed among PC patients. In both groups the death rate was low (3%, n = 1/34 for PC and 2%, n = 1/41 for DC); one patient died due to PJP and one died due to lymphoma.\\nThe incidence of LD (p = 0.020), CD4 count < 350 cells/μL (p = 0.037) and < 200 cells /μL (p = 0.022) were statistically significantly associated with the ongoing COVID pandemic. Association with transmission risk was borderline significant (p = 0.055).\",\n \"To our knowledge this is the first real‐life cohort presenting novel data on the impact of COVID‐19 in HIV late diagnosis in Germany.\\nEven with easy accessibility to testing facilities and the prompt treatment of newly diagnosed HIV patients, 40–55% of patients are still diagnosed with late diagnosis in Germany [5, 6]. Other European cohorts have shown similar LD rates between 42% and 58% [7, 8, 9, 10]. In the COHERE cohort, being heterosexual, originating from southern Europe or being of African descent were found to be associated with the highest risk of LD [10]. Older age has also been associated with an increased risk of LD [5]. It is an urgent goal to diagnose HIV patients at an early stage as it is a known fact that LD leads to increased morbidity and mortality as well as increased health costs. The UNAIDS 95‐95‐95 goals are at increasing risk of being missed due to the ongoing focus on COVID‐19 and the resulting missed opportunities of treating HIV patients. As reported in the UNAIDS HIV Services Tracking, a large, sustained decrease in HIV testing in the reporting countries was observed [11].\\nWith the beginning of the COVID‐19 pandemic in Germany in February 2020, we found the rate of LD to be rising at our infectious disease department. The overall rate of LD during the ongoing pandemic has shown a 40% increase over the pre‐pandemic time‐frame, resulting in a 3.4‐fold higher odds ratio (95% confidence interval: 1.18–9.83). This is most likely due to the numerous impacts of the ongoing COVID‐9 pandemic in Germany. Beginning with increasing numbers of SARS‐CoV‐2 infections in spring 2020, the fear of acquiring SARS‐CoV‐2 infection rose steadily, which led to an overall decreased consultation rate by, and access to, physicians. Therefore, possible symptoms of progressed HIV infection might have been neglected or their investigation postponed by the individual, or been misinterpreted due to the main focus on COVID‐19.\\nAccess to low‐threshold testing was impeded by down‐regulation of community service testing at checkpoints and municipal health centres. As a result, two main areas for detecting undiagnosed HIV infection (whether with symptoms or asymptomatic) were disrupted.\\nConcerning risk of transmission, we found a decreasing rate of new HIV infections in MSM which was also observed in an Italian cohort during the ongoing COVID‐19 pandemic [12]. This could be due to reduced contacts during lockdown as well as the fewer possibilities for testing mentioned earlier, which is more widespread and implemented in this risk group. Previous cohorts have shown a decrease in late diagnosis in MSM over time, which might be due to the implementation of pre‐exposure prophylaxis [10]. In our study we found an increase in LD among heterosexuals. Moreover, transmission via heterosexual contact has been described earlier as an associated risk factor for LD [11]. Indeed, this group might be more affected by having less information about testing strategies as well as by a lack of awareness regarding their individual risk of infection and could therefore have been even more impeded during the ongoing COVID‐19 pandemic.\\nRegarding the institution in which HIV infection was diagnosed, we found a shift towards diagnosis in hospital due to mandatory treatment, whereas diagnosis before COVID‐19 was most often performed in specialist practices. However, due to the high rate of LD in our cohort, this finding has to be interpreted with caution. This reflects the fact that ongoing LD during the COVID‐19 pandemic might be due to the reduced availability of low‐threshold testing and reduced slots at general practitioners. Moreover, LD, with its concomitant laboratory and clinical abnormalities, requires special awareness, underlining the fact that all physicians, regardless of their speciality, need to be aware of HIV‐associated symptoms and disease in order to determine the HIV diagnosis.\\nA low CD4 count reflects the impaired immune function and increased risk of developing AIDS‐defining diseases, which all lead to increased morbidity and mortality. In our study, CD4 counts < 350 cells/μL and < 200 cells/μL were statistically significantly associated with the ongoing COVID‐19 pandemic, underlining the fact that we need to put all our efforts towards diagnosing patients at an earlier stage in order to prevent severe immune dysfunction and subsequent AIDS‐defining diseases.\\nAt the same time, we found an almost doubled rate of AIDS‐defining diseases during the ongoing COVID‐19 pandemic. PJP, which is the most prevalent AIDS‐defining disease, was found most often in our study. It is important to note that more than half of patients with PJP are so severely ill that they require intensive care treatment; 40% (n = 2/5) of DC patients with PJP who required intensive care treatment had to undergo mechanical ventilation via intubation, as compared with 25% (n = 1/4) of PC patients. The overall increased rate of AIDS‐defining diseases raises morbidity and mortality, which also increases the cost of healthcare. Moreover, persistent organ damage or development of renal insufficiency after recovering from PJP or other AIDS‐defining diseases has to be taken into consideration.\\nInterestingly, we diagnosed several patients with AIDS‐related neurological disorders, whereas none had been observed in the comparable pre‐COVID cohort. Since implementation of cART, incidence of HIV‐associated neurological diseases has been decreasing, and thus patients with LD are still at an elevated risk [13]. In particular, AIDS‐related neurological disorders cause significant and, most often, permanent harm to the patient and therefore early diagnosis is of vital importance. Both of our patients with cerebral toxoplasmosis still have tremendous neurological deficits. One patient who presented with hemiplegia due to cerebral toxoplasmosis and severely spread Kaposi sarcoma even went blind due to cytomegalovirus‐associated retinitis.\\nOur study has several limitations. First, it is a monocentric study, and second it only reflects data from two time slots which are close together. The overall higher rate of LD might be overestimated due to coincidence and the historical fact of the increased amount of LD in our hospital, as it is well known for its special infectious disease department. Thus, this study reflects a real‐life cohort being affected by the ongoing COVID‐19 pandemic.\\nIn conclusion we found an increased rate of LD in our cohort during the persisting COVID‐19 pandemic despite stable annual numbers of new HIV diagnoses, which underlines the barriers to HIV testing as well as general care as a result of the focus on COVID‐19. The outcome is more AIDS‐defining diseases and higher hospitalization rates. Maintaining HIV testing opportunities and access to treatment during a pandemic is crucial if it is not to impede WHO elimination goals and we are to prevent an increase in AIDS‐related morbidity and mortality.\",\n \"KvB designed the study and concept. Sample collection was performed by all co‐authors. Data analysis was performed by KvB. The manuscript was written by KvB and CB with full input from all co‐authors.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,"results","discussion",null],"string":"[\n null,\n \"methods\",\n null,\n \"results\",\n \"discussion\",\n null\n]"},"keywords":{"kind":"list like","value":["CD4 count","COVID‐19","HIV","late diagnosis"],"string":"[\n \"CD4 count\",\n \"COVID‐19\",\n \"HIV\",\n \"late diagnosis\"\n]"}}},{"rowIdx":358,"cells":{"title":{"kind":"string","value":"Clinical evaluation of a novel molecular diagnosis kit for detecting Helicobacter pylori and clarithromycin-resistant using intragastric fluid."},"pmid":{"kind":"string","value":"36263754"},"background_abstract":{"kind":"string","value":"Although there are many Helicobacter pylori (H. pylori) diagnostic methods, the culture and antibiotic susceptibility test is an important method for selecting the most effective H. pylori eradication regimen. However, this diagnostic method is complicated and takes several days; therefore, the development of a rapid and simple diagnostic method is required. Eradication failure due to clarithromycin (CAM) resistance should also be considered. In this study, we report the clinical evaluation of point-of-care testing (POCT) kit using intragastric fluid, a novel kit for detecting H. pylori and CAM resistance."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"The study participants were 143 patients suspected of H. pylori infection and had an endoscopic examination. The novel diagnostic kit diagnosed H. pylori infection and CAM resistance-associated mutation using intragastric fluid. To diagnose H. pylori infection, the relationship between the diagnostic kit and conventional diagnostic methods (urea breath test, stool antigen test, culture test, and real-time polymerase chain reaction [PCR]) was evaluated. For CAM resistance-associated mutation detection, the concordance between the diagnostic kit and antibiotic susceptibility test was evaluated."},"methods_abstract_label":{"kind":"string","value":"MATERIALS AND METHODS"},"results_abstract":{"kind":"string","value":"The diagnosis of H. pylori infection with the novel molecular diagnostic kit using intragastric fluid showed significant relationship with conventional diagnostic methods. Especially when the culture was control, the sensitivity was 100% (67/67), the specificity was 95.9% (71/74), and the overall concordance was 97.9% (138/141). The detection of CAM resistance-associated mutations had a concordance rate of 97.0% (65/67) when compared with the antibiotic susceptibility test."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The H. pylori molecular POCT kit uses intragastric fluid as a sample and can diagnose H. pylori infection and detect CAM resistance-associated mutations within an hour. This novel kit is expected to prove useful in selecting the most effective eradication regimen for H. pylori."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Helicobacter pylori","Clarithromycin","Helicobacter Infections","Drug Resistance, Bacterial","Anti-Bacterial Agents","Microbial Sensitivity Tests"],"string":"[\n \"Humans\",\n \"Helicobacter pylori\",\n \"Clarithromycin\",\n \"Helicobacter Infections\",\n \"Drug Resistance, Bacterial\",\n \"Anti-Bacterial Agents\",\n \"Microbial Sensitivity Tests\"\n]"},"pmcid":{"kind":"string","value":"9788249"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis [SUBSECTION] Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\nResults from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\n[SUBTITLE] Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene [SUBSECTION] The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\nThe mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\n[SUBTITLE] Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\n23S rRNA by pyrosequencing [SUBSECTION] Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\nQuantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\n[SUBTITLE] Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\n [SUBSECTION] “Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test.\n“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Patient enrollment and Helicobacter pylori conventional testing","Sample collection of gastric mucosal biopsy and intragastric fluid","\nHelicobacter pylori antibiotic susceptibility test","Smart Gene assay for detection of Helicobacter pylori\nDNA and CAM‐resistant mutation","Real‐time PCR of Helicobacter pylori\n16S rRNA\n","Quantitative analysis of Helicobacter pylori\nCAM‐resistant mutation ratio by pyrosequencing","Statistical analysis","Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis","Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene","Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\n23S rRNA by pyrosequencing","Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\n","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION"],"string":"[\n \"INTRODUCTION\",\n \"Patient enrollment and Helicobacter pylori conventional testing\",\n \"Sample collection of gastric mucosal biopsy and intragastric fluid\",\n \"\\nHelicobacter pylori antibiotic susceptibility test\",\n \"Smart Gene assay for detection of Helicobacter pylori\\nDNA and CAM‐resistant mutation\",\n \"Real‐time PCR of Helicobacter pylori\\n16S rRNA\\n\",\n \"Quantitative analysis of Helicobacter pylori\\nCAM‐resistant mutation ratio by pyrosequencing\",\n \"Statistical analysis\",\n \"Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis\",\n \"Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene\",\n \"Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\\n23S rRNA by pyrosequencing\",\n \"Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\\n\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\"\n]"},"other_sections_texts":{"kind":"list like","value":["\nHelicobacter pylori (H. pylori) infection causes various upper gastrointestinal disorders such as atrophic gastritis, gastric ulcer, duodenal ulcer, gastric cancer, gastric mucosa associated lymphoid tissue (MALT) lymphoma, and gastric hyperplastic polyp associated with chronic inflammation of the gastric mucosa.\n1\n\nH. pylori infection is associated with gastric cancer as an important cause. The International Agency for Research on Cancer, the cancer research agency of the World Health Organization, strongly recommends designing a global preventive method for gastric cancer, including H. pylori eradication therapy.\n2\n, \n3\n\n\nIn Japan, “H. pylori‐infected gastritis” has been included in the list of diseases covered by insurances since 2013. Consequently, eradication therapy for chronic gastritis is compensated by the national health insurance scheme. Interestingly, there has been a decrease in gastric cancer deaths since the start of the insurance coverage.\n4\n\n\nThe Japanese Society for Helicobacter Research guidelines recommend H. pylori eradication therapy as the primary preventive method for gastric cancer after antibiotic susceptibility test. Furthermore, the guidelines recommend the use of the therapy in two or three drug combinations to maximize the bacteria eradication rate because of possible eradication therapy failure.\n1\n The failure of H. pylori eradication therapy is largely due to the resistance to clarithromycin (CAM), which is the primary drug for primary eradication. CAM resistance is caused by a gene mutation at positions 2142 and 2143 of the 23S rRNA gene Domain V region of H. pylori.\n5\n We previously reported the applications of gastric wash sample (intragastric fluid), such as the detection of CAM resistance‐associated gene mutation of H. pylori, for selecting eradication therapy and quantitative analysis of H. pylori genotype via pyrosequencing analysis.\n6\n This detection method using gastric wash sample (intragastric fluid) results in non‐invasive and high safety compared with the collect gastric mucosal biopsy for culture and susceptibility testing.\n6\n, \n7\n In addition, the results of antibiotic susceptibility testing take about 1 week; therefore, the method is too time‐consuming and impractical in clinical practice. Therefore, physicians need a new, time‐saving test to determine antibiotic susceptibility.\nSmart Gene™ (MIZUHO MEDY Co., Ltd.) was developed on the concept of point‐of‐care testing (POCT), which can automatically perform nucleic acid extraction, amplification, and detection. The detection of Mycoplasma pneumoniae and coronavirus by Smart Gene™ with pharyngeal swabs shows high reproducibility and useability and allows quick patient triage.\n8\n, \n9\n, \n10\n, \n11\n In addition, the detection of H. pylori by Smart Gene™ with stool proved to be an effective non‐invasive test.\n12\n The “H. pylori gene detection POCT reagent,” a reagent for the fully automatic Smart Gene™, allows for the automatic detection of H. pylori and CAM resistance‐associated mutations based on the Qprobe method.\n13\n, \n14\n, \n15\n\n\nOur goal was the fast and accurate diagnosis of H. pylori and detection of CAM resistance for the purpose of determining a precise eradication therapy. Therefore, we evaluated the clinical performance of the “H. pylori molecular POCT kit” in diagnosing H. pylori infection and detecting CAM‐resistant mutations. In this paper, we report new fast and accurate clinical technique for the diagnosis H. pylori infection in as fast as about an hour with the kit by Smart Gene™.","To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test","To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\n6\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™","To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\n16\n, \n17\n\n","To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\n10\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).","To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\n18\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.","To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\n6\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.","Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).","Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.","The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.","Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin","“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test.","MT performed the experiments and analyses and drafted the manuscript. YW performed the experiments. RO performed pyrosequencing. RW, MH, and KK performed the experiments. HY and FI supervised the experiments. MK supervised the entire project.","MIZUHO MEDY Co., Ltd provided grants and H. pylori molecular POCT kit for this study, but played no role in the study design, data collection and interpretation, or in the decision to submit the work for publication."],"string":"[\n \"\\nHelicobacter pylori (H. pylori) infection causes various upper gastrointestinal disorders such as atrophic gastritis, gastric ulcer, duodenal ulcer, gastric cancer, gastric mucosa associated lymphoid tissue (MALT) lymphoma, and gastric hyperplastic polyp associated with chronic inflammation of the gastric mucosa.\\n1\\n\\nH. pylori infection is associated with gastric cancer as an important cause. The International Agency for Research on Cancer, the cancer research agency of the World Health Organization, strongly recommends designing a global preventive method for gastric cancer, including H. pylori eradication therapy.\\n2\\n, \\n3\\n\\n\\nIn Japan, “H. pylori‐infected gastritis” has been included in the list of diseases covered by insurances since 2013. Consequently, eradication therapy for chronic gastritis is compensated by the national health insurance scheme. Interestingly, there has been a decrease in gastric cancer deaths since the start of the insurance coverage.\\n4\\n\\n\\nThe Japanese Society for Helicobacter Research guidelines recommend H. pylori eradication therapy as the primary preventive method for gastric cancer after antibiotic susceptibility test. Furthermore, the guidelines recommend the use of the therapy in two or three drug combinations to maximize the bacteria eradication rate because of possible eradication therapy failure.\\n1\\n The failure of H. pylori eradication therapy is largely due to the resistance to clarithromycin (CAM), which is the primary drug for primary eradication. CAM resistance is caused by a gene mutation at positions 2142 and 2143 of the 23S rRNA gene Domain V region of H. pylori.\\n5\\n We previously reported the applications of gastric wash sample (intragastric fluid), such as the detection of CAM resistance‐associated gene mutation of H. pylori, for selecting eradication therapy and quantitative analysis of H. pylori genotype via pyrosequencing analysis.\\n6\\n This detection method using gastric wash sample (intragastric fluid) results in non‐invasive and high safety compared with the collect gastric mucosal biopsy for culture and susceptibility testing.\\n6\\n, \\n7\\n In addition, the results of antibiotic susceptibility testing take about 1 week; therefore, the method is too time‐consuming and impractical in clinical practice. Therefore, physicians need a new, time‐saving test to determine antibiotic susceptibility.\\nSmart Gene™ (MIZUHO MEDY Co., Ltd.) was developed on the concept of point‐of‐care testing (POCT), which can automatically perform nucleic acid extraction, amplification, and detection. The detection of Mycoplasma pneumoniae and coronavirus by Smart Gene™ with pharyngeal swabs shows high reproducibility and useability and allows quick patient triage.\\n8\\n, \\n9\\n, \\n10\\n, \\n11\\n In addition, the detection of H. pylori by Smart Gene™ with stool proved to be an effective non‐invasive test.\\n12\\n The “H. pylori gene detection POCT reagent,” a reagent for the fully automatic Smart Gene™, allows for the automatic detection of H. pylori and CAM resistance‐associated mutations based on the Qprobe method.\\n13\\n, \\n14\\n, \\n15\\n\\n\\nOur goal was the fast and accurate diagnosis of H. pylori and detection of CAM resistance for the purpose of determining a precise eradication therapy. Therefore, we evaluated the clinical performance of the “H. pylori molecular POCT kit” in diagnosing H. pylori infection and detecting CAM‐resistant mutations. In this paper, we report new fast and accurate clinical technique for the diagnosis H. pylori infection in as fast as about an hour with the kit by Smart Gene™.\",\n \"To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\",\n \"To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\\n6\\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\",\n \"To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\\n16\\n, \\n17\\n\\n\",\n \"To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\\n10\\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\",\n \"To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\\n18\\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\",\n \"To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\\n6\\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\",\n \"Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).\",\n \"Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\",\n \"The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\\nList of the mismatch cases between Smart Gene™ assay and control methods\\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\",\n \"Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\",\n \"“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\\nComparison of CAM resistance determined by Smart Gene™ assay and AST\\nAbbreviation: AST, Antibiotic susceptibility test.\",\n \"MT performed the experiments and analyses and drafted the manuscript. YW performed the experiments. RO performed pyrosequencing. RW, MH, and KK performed the experiments. HY and FI supervised the experiments. MK supervised the entire project.\",\n \"MIZUHO MEDY Co., Ltd provided grants and H. pylori molecular POCT kit for this study, but played no role in the study design, data collection and interpretation, or in the decision to submit the work for publication.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","MATERIAL AND METHODS","Patient enrollment and Helicobacter pylori conventional testing","Sample collection of gastric mucosal biopsy and intragastric fluid","\nHelicobacter pylori antibiotic susceptibility test","Smart Gene assay for detection of Helicobacter pylori\nDNA and CAM‐resistant mutation","Real‐time PCR of Helicobacter pylori\n16S rRNA\n","Quantitative analysis of Helicobacter pylori\nCAM‐resistant mutation ratio by pyrosequencing","Statistical analysis","RESULTS","Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis","Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene","Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\n23S rRNA by pyrosequencing","Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\n","DISCUSSION","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","CONFLICT OF INTEREST"],"string":"[\n \"INTRODUCTION\",\n \"MATERIAL AND METHODS\",\n \"Patient enrollment and Helicobacter pylori conventional testing\",\n \"Sample collection of gastric mucosal biopsy and intragastric fluid\",\n \"\\nHelicobacter pylori antibiotic susceptibility test\",\n \"Smart Gene assay for detection of Helicobacter pylori\\nDNA and CAM‐resistant mutation\",\n \"Real‐time PCR of Helicobacter pylori\\n16S rRNA\\n\",\n \"Quantitative analysis of Helicobacter pylori\\nCAM‐resistant mutation ratio by pyrosequencing\",\n \"Statistical analysis\",\n \"RESULTS\",\n \"Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis\",\n \"Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene\",\n \"Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\\n23S rRNA by pyrosequencing\",\n \"Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\\n\",\n \"DISCUSSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"CONFLICT OF INTEREST\"\n]"},"all_sections_texts":{"kind":"list like","value":["\nHelicobacter pylori (H. pylori) infection causes various upper gastrointestinal disorders such as atrophic gastritis, gastric ulcer, duodenal ulcer, gastric cancer, gastric mucosa associated lymphoid tissue (MALT) lymphoma, and gastric hyperplastic polyp associated with chronic inflammation of the gastric mucosa.\n1\n\nH. pylori infection is associated with gastric cancer as an important cause. The International Agency for Research on Cancer, the cancer research agency of the World Health Organization, strongly recommends designing a global preventive method for gastric cancer, including H. pylori eradication therapy.\n2\n, \n3\n\n\nIn Japan, “H. pylori‐infected gastritis” has been included in the list of diseases covered by insurances since 2013. Consequently, eradication therapy for chronic gastritis is compensated by the national health insurance scheme. Interestingly, there has been a decrease in gastric cancer deaths since the start of the insurance coverage.\n4\n\n\nThe Japanese Society for Helicobacter Research guidelines recommend H. pylori eradication therapy as the primary preventive method for gastric cancer after antibiotic susceptibility test. Furthermore, the guidelines recommend the use of the therapy in two or three drug combinations to maximize the bacteria eradication rate because of possible eradication therapy failure.\n1\n The failure of H. pylori eradication therapy is largely due to the resistance to clarithromycin (CAM), which is the primary drug for primary eradication. CAM resistance is caused by a gene mutation at positions 2142 and 2143 of the 23S rRNA gene Domain V region of H. pylori.\n5\n We previously reported the applications of gastric wash sample (intragastric fluid), such as the detection of CAM resistance‐associated gene mutation of H. pylori, for selecting eradication therapy and quantitative analysis of H. pylori genotype via pyrosequencing analysis.\n6\n This detection method using gastric wash sample (intragastric fluid) results in non‐invasive and high safety compared with the collect gastric mucosal biopsy for culture and susceptibility testing.\n6\n, \n7\n In addition, the results of antibiotic susceptibility testing take about 1 week; therefore, the method is too time‐consuming and impractical in clinical practice. Therefore, physicians need a new, time‐saving test to determine antibiotic susceptibility.\nSmart Gene™ (MIZUHO MEDY Co., Ltd.) was developed on the concept of point‐of‐care testing (POCT), which can automatically perform nucleic acid extraction, amplification, and detection. The detection of Mycoplasma pneumoniae and coronavirus by Smart Gene™ with pharyngeal swabs shows high reproducibility and useability and allows quick patient triage.\n8\n, \n9\n, \n10\n, \n11\n In addition, the detection of H. pylori by Smart Gene™ with stool proved to be an effective non‐invasive test.\n12\n The “H. pylori gene detection POCT reagent,” a reagent for the fully automatic Smart Gene™, allows for the automatic detection of H. pylori and CAM resistance‐associated mutations based on the Qprobe method.\n13\n, \n14\n, \n15\n\n\nOur goal was the fast and accurate diagnosis of H. pylori and detection of CAM resistance for the purpose of determining a precise eradication therapy. Therefore, we evaluated the clinical performance of the “H. pylori molecular POCT kit” in diagnosing H. pylori infection and detecting CAM‐resistant mutations. In this paper, we report new fast and accurate clinical technique for the diagnosis H. pylori infection in as fast as about an hour with the kit by Smart Gene™.","[SUBTITLE] Patient enrollment and Helicobacter pylori conventional testing [SUBSECTION] To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\nTo collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\n[SUBTITLE] Sample collection of gastric mucosal biopsy and intragastric fluid [SUBSECTION] To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\n6\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\nTo diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\n6\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\n[SUBTITLE] \nHelicobacter pylori antibiotic susceptibility test [SUBSECTION] To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\n16\n, \n17\n\n\nTo determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\n16\n, \n17\n\n\n[SUBTITLE] Smart Gene assay for detection of Helicobacter pylori\nDNA and CAM‐resistant mutation [SUBSECTION] To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\n10\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\nTo evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\n10\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\n[SUBTITLE] Real‐time PCR of Helicobacter pylori\n16S rRNA\n [SUBSECTION] To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\n18\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\nTo compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\n18\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\n[SUBTITLE] Quantitative analysis of Helicobacter pylori\nCAM‐resistant mutation ratio by pyrosequencing [SUBSECTION] To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\n6\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\nTo assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\n6\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\n[SUBTITLE] Statistical analysis [SUBSECTION] Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).\nConcordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).","To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test","To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\n6\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™","To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\n16\n, \n17\n\n","To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\n10\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).","To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\n18\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.","To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\n6\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.","Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).","[SUBTITLE] Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis [SUBSECTION] Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\nResults from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\n[SUBTITLE] Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene [SUBSECTION] The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\nThe mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\n[SUBTITLE] Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\n23S rRNA by pyrosequencing [SUBSECTION] Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\nQuantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\n[SUBTITLE] Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\n [SUBSECTION] “Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test.\n“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test.","Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.","The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\nList of the mismatch cases between Smart Gene™ assay and control methods\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.","Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin","“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\nComparison of CAM resistance determined by Smart Gene™ assay and AST\nAbbreviation: AST, Antibiotic susceptibility test.","“Helicobacter pylori molecular POCT kit” by Smart Gene™ allows physicians to safely collect specimens and obtain diagnostic results on time. The simplicity of the kit allows for fast and accurate diagnosis of CAM resistance in determining a patient's precise eradication therapy. Consequently, its wider application will lead to more patients being tested for their H. pylori infection status.\nThe effectiveness of the “H. pylori molecular POCT kit” in detecting a CAM resistance‐associated mutation rate at a high concordance rate as an antibiotic susceptibility test proves beneficial in daily clinical practice. This novel kit can detect gene mutation at positions 2142 and 2143 of the 23S rRNA gene domain V region of H. pylori, precisely indicating CAM resistance. Furthermore, the “H. pylori molecular POCT kit” can detect CAM resistance‐associated mutations at a mutation rate of 20% or more when compared to pyrosequencing analysis. Results obtained by “H. pylori molecular POCT kit” in detecting CAM resistance is a great indication of patients' eradication results.\nUsing “H. pylori molecular POCT kit” for detecting CAM‐resistant mutation can significantly improve time‐efficiency in planning a specific treatment plan for patients. Due to its simplicity and speed, compared with other existing antibiotic susceptibility tests, physicians can speed up the process of selecting the appropriate antibiotic for the eradication of H. pylori, allowing patients to receive treatment as early as possible. As a result of the quick results provided by “H. pylori molecular POCT kit,” physicians can provide individualized treatment for patients with CAM resistance, such as skipping CAM‐based therapy and jumping into non‐CAM‐based therapy. Consequently, this will save physicians a significant amount of time they would have wasted on waiting for results and trying out ineffective treatments.\nAdditionally, “H. pylori molecular POCT kit” is safer for patients than conventional tests because it requires no additional invasive procedure during endoscopic examination and offers painless procedures. Patients only need to provide their intragastric fluid as a test specimen and do not have to go through gastric mucosal biopsy, which could cause complications, including excessive bleeding. Furthermore, because collecting intragastric fluid requires very little preparation time, patients do not have to suffer any pain, which they are likely to experience with longer preparation time. This significantly lower the risk of patient's vital transition.\nThe ease‐of‐use and reliability of results with the “H. pylori molecular POCT kit” could lead to the consolidation of all the different H. pylori diagnostic tests currently in use. Although the occurrence of false positives and negatives need to be further investigated, the kit has a high likelihood of replacing existing diagnostic tests. In this study, three cases (Case 4, 5, 13) tested positive for only PCR assumed to be the result of insufficient H. pylori amount.\n“H. pylori molecular POCT kit” is so simple and so capable of accurately and safely diagnosing H. pylori infection that it could be a great alternative diagnostic method in near future. Because the kit has a wide application and allows for the simultaneous testing of H. pylori infection and CAM resistance statuses, physicians can save time on eradication failure caused by CAM resistance.\nOne limitation of our study is the sample size. Although our sample size was sufficient enough to give us acceptable data, to increase the accuracy of the kit, we would like to enroll and study more people diagnosed with H. pylori infection. Because participating the study and eradicating H. pylori only lower participants' risk of developing gastric cancer, this study is also a part of eradicating overall gastric cancer itself.","MT performed the experiments and analyses and drafted the manuscript. YW performed the experiments. RO performed pyrosequencing. RW, MH, and KK performed the experiments. HY and FI supervised the experiments. MK supervised the entire project.","MIZUHO MEDY Co., Ltd provided grants and H. pylori molecular POCT kit for this study, but played no role in the study design, data collection and interpretation, or in the decision to submit the work for publication.","The authors declare no conflict of interests associated with this study."],"string":"[\n \"\\nHelicobacter pylori (H. pylori) infection causes various upper gastrointestinal disorders such as atrophic gastritis, gastric ulcer, duodenal ulcer, gastric cancer, gastric mucosa associated lymphoid tissue (MALT) lymphoma, and gastric hyperplastic polyp associated with chronic inflammation of the gastric mucosa.\\n1\\n\\nH. pylori infection is associated with gastric cancer as an important cause. The International Agency for Research on Cancer, the cancer research agency of the World Health Organization, strongly recommends designing a global preventive method for gastric cancer, including H. pylori eradication therapy.\\n2\\n, \\n3\\n\\n\\nIn Japan, “H. pylori‐infected gastritis” has been included in the list of diseases covered by insurances since 2013. Consequently, eradication therapy for chronic gastritis is compensated by the national health insurance scheme. Interestingly, there has been a decrease in gastric cancer deaths since the start of the insurance coverage.\\n4\\n\\n\\nThe Japanese Society for Helicobacter Research guidelines recommend H. pylori eradication therapy as the primary preventive method for gastric cancer after antibiotic susceptibility test. Furthermore, the guidelines recommend the use of the therapy in two or three drug combinations to maximize the bacteria eradication rate because of possible eradication therapy failure.\\n1\\n The failure of H. pylori eradication therapy is largely due to the resistance to clarithromycin (CAM), which is the primary drug for primary eradication. CAM resistance is caused by a gene mutation at positions 2142 and 2143 of the 23S rRNA gene Domain V region of H. pylori.\\n5\\n We previously reported the applications of gastric wash sample (intragastric fluid), such as the detection of CAM resistance‐associated gene mutation of H. pylori, for selecting eradication therapy and quantitative analysis of H. pylori genotype via pyrosequencing analysis.\\n6\\n This detection method using gastric wash sample (intragastric fluid) results in non‐invasive and high safety compared with the collect gastric mucosal biopsy for culture and susceptibility testing.\\n6\\n, \\n7\\n In addition, the results of antibiotic susceptibility testing take about 1 week; therefore, the method is too time‐consuming and impractical in clinical practice. Therefore, physicians need a new, time‐saving test to determine antibiotic susceptibility.\\nSmart Gene™ (MIZUHO MEDY Co., Ltd.) was developed on the concept of point‐of‐care testing (POCT), which can automatically perform nucleic acid extraction, amplification, and detection. The detection of Mycoplasma pneumoniae and coronavirus by Smart Gene™ with pharyngeal swabs shows high reproducibility and useability and allows quick patient triage.\\n8\\n, \\n9\\n, \\n10\\n, \\n11\\n In addition, the detection of H. pylori by Smart Gene™ with stool proved to be an effective non‐invasive test.\\n12\\n The “H. pylori gene detection POCT reagent,” a reagent for the fully automatic Smart Gene™, allows for the automatic detection of H. pylori and CAM resistance‐associated mutations based on the Qprobe method.\\n13\\n, \\n14\\n, \\n15\\n\\n\\nOur goal was the fast and accurate diagnosis of H. pylori and detection of CAM resistance for the purpose of determining a precise eradication therapy. Therefore, we evaluated the clinical performance of the “H. pylori molecular POCT kit” in diagnosing H. pylori infection and detecting CAM‐resistant mutations. In this paper, we report new fast and accurate clinical technique for the diagnosis H. pylori infection in as fast as about an hour with the kit by Smart Gene™.\",\n \"[SUBTITLE] Patient enrollment and Helicobacter pylori conventional testing [SUBSECTION] To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\\nTo collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\\n[SUBTITLE] Sample collection of gastric mucosal biopsy and intragastric fluid [SUBSECTION] To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\\n6\\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\\nTo diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\\n6\\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\\n[SUBTITLE] \\nHelicobacter pylori antibiotic susceptibility test [SUBSECTION] To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\\n16\\n, \\n17\\n\\n\\nTo determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\\n16\\n, \\n17\\n\\n\\n[SUBTITLE] Smart Gene assay for detection of Helicobacter pylori\\nDNA and CAM‐resistant mutation [SUBSECTION] To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\\n10\\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\\nTo evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\\n10\\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\\n[SUBTITLE] Real‐time PCR of Helicobacter pylori\\n16S rRNA\\n [SUBSECTION] To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\\n18\\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\\nTo compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\\n18\\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\\n[SUBTITLE] Quantitative analysis of Helicobacter pylori\\nCAM‐resistant mutation ratio by pyrosequencing [SUBSECTION] To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\\n6\\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\\nTo assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\\n6\\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\\n[SUBTITLE] Statistical analysis [SUBSECTION] Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).\\nConcordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).\",\n \"To collect samples through esophagogastroduodenoscopy (EGD) for this study, we assessed the eligibility of 151 patients suspected of H. pylori infection at the National Hospital Organization Hakodate Hospital and the Kawasaki Rinko Hospital from December 2019 to March 2021. However, a total 8 patients were excluded from the study because they withdraw their consent (n = 2) or did not undergo EGD (n = 6); therefore, 143 patients were enrolled in the study (Figure 1). In addition to EGD, selected patients underwent a urea breath test (UBT, Urea breath test, UBIT tablet, 100 mg; Otsuka Pharmaceutical Co., Ltd.) stool antigen test (SAT, Stool antigen test, Testmate rapid pylori antigen; Wakamoto Pharmaceutical Co., Ltd) and H. pylori culture test (BML Co., Ltd.) as part of the conventional tests required for the diagnosis of H. pylori infection. From the result of the tests, 70 patients were confirmed to be H. pylori‐positive and 73 patients were confirmed to be H. pylori‐negative.\\nThe flowchart of participants. We finally enrolled 143 patients for whom we performed endoscopic examination to detect Helicobacter pylori infection and clarithromycin‐resistant mutation by conventional testing, pyrosequencing, and H. pylori gene measurement. AST, Antibiotic susceptibility test; EGD, esophagogastroduodenoscopy; PCR, Polymerase chain reaction; SAT, Stool antigen test; UBT, Urea breath test\",\n \"To diagnose H. pylori infection and antibiotic susceptibility for CAM, we collected the patients' intragastric fluid and performed gastric mucosal biopsy as part of the endoscopic examination procedure, which has been previously reported.\\n6\\n Patient drank 100 ml of water (including 80 mg dimethylpolysiloxane, 1 g sodium bicarbonate, and 20,000 units of pronase) 10 min prior to their endoscopy. We collected a sample tissue of approximately 5 mm under endoscopic observation using biopsy forceps. Patients' intragastric fluid was collected directly to a sample container (MD‐33050, SB‐Kawasumi Laboratories, Inc.) through the connected suction base of the endoscope (Figure 2A).\\n(A) Collection of intragastric fluid using sample container. (B) Assay workflow for Helicobacter pylori molecular POCT kit with Smart Gene™\",\n \"To determine H. pylori susceptibility, we performed antibiotic susceptibility testing of CAM using gastric mucosal biopsy as a conventional diagnosing method. The susceptibility tests were conducted by the clinical testing contractor BML Co., Ltd. In detail, gastric biopsies were cultured on Nissui Plate Helicobacter agar (Nissui Pharmaceutical Co., Ltd.) under microaerophilic conditions (5% O2, 15% CO2, and 80% N2) for 4 days at 37°C. Then, the isolated 3 colonies were enriched on 5% sheep blood agar (Nippon Becton Dickinson Co., Ltd.). The enriched colonies were prepared to a McF1.0 bacterial solution, and 25 μl bacterial solution was added to 6 ml of 10% horse serum‐added Mueller Hinton broth (Nikken Biological Co., Ltd.) to prepare a sample. Antibiotic susceptibility testing of CAM was performed using the dry plate “Eiken” (Eiken Chemical Co., Ltd.). The antibiotic susceptibility of CAM was determined using 1 μg/ml, minimal inhibitory concentration (MIC) breakpoint, recommended by the Japan Society of Chemotherapy and the Clinical & Laboratory Standards Institute (CLSI).\\n16\\n, \\n17\\n\\n\",\n \"To evaluate the effectiveness of the novel “H. pylori molecular POCT kit” in detecting H. pylori and CAM‐resistant mutations using Smart Gene™ (Mizuho Medy Co., Ltd.), we used intragastric fluid as a sample. We simply dropped intragastric fluid suspended into the test cartridge “H. pylori molecular POCT kit.” Smart Gene™ detects the H. pylori DNA and CAM‐resistant mutations at positions 2142 and 2143 of the 23S rRNA gene based on polymerase chain reaction (PCR) and QProbe. The principle of detecting gene mutations by QProbe is based on previous report that demonstrated use of the macrolide‐resistant Mycoplasma pneumoniae (MRMp) diagnostic kit.\\n10\\n The use of Smart Gene™ allows for the simultaneous diagnosis of H. pylori infection and detection CAM resistance‐associated mutation. In our study, the Smart Gene™ automatically processed the samples and generated the results within an hour. In addition, because we utilized intragastric fluid, there was no additional invasive procedure carried out on study participants at the time of endoscopy (Figure 2B).\",\n \"To compare the outcomes of different testing methods, we performed a molecular diagnostic real‐time PCR targeting the H. pylori 16S rRNA gene as our control.\\n18\\n We extracted DNA from 200 μl intragastric fluid with the QIAamp DNA Mini Kit (QIAGEN GmbH) to obtain 150 μl purified DNA. The forward primer CGC‐TAA‐GAG‐ATC‐AGC‐CTA‐TGT‐CC and the reverse primer CCG‐TGT‐CTC‐AGT‐TCC‐AGT‐GTG‐T were used for real‐time PCR. Real‐time PCR was performed using the Thermal Cycler Dice Real Time System III instrument (Takara Bio) using TB Green Premix Dimer Eraser Perfect Real Time reagent (Takara Bio) under the following PCR conditions: preheating at 95°C for 30 s, and 50 cycles at 95°C for 5 s and at 55°C for 30 s and at 72°C for 30 s. The PCR amplicon was confirmed by melting curve analysis.\",\n \"To assess and determine CAM resistance, we quantitatively analyzed mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene via pyrosequencing, adopting previously reported procedures.\\n6\\n One milliliters intragastric fluid was centrifuged at 1200 g for 15 min and 800 μl supernatant was decanted. 200 μl purified DNA was extracted from residue sample with the QIAamp DNA Mini Kit (QIAGEN GmbH). Thereafter, the H. pylori 23S rRNA gene was amplified and biotinylated by nested PCR. In the first reaction, a 255‐bp fragment was amplified with the forward primer ACG‐AGA‐TGG‐GAG‐CTG‐TCT‐CAA‐CC and the reverse primer AGC‐ATT‐GTC‐CTG‐CCT‐GTG‐GAT‐AAC. The amplified fragments were, thereafter, used as a template in the second reaction to amplify a 90‐bp fragment with the forward primer GAG‐GTG‐AAA‐ATT‐CCT‐CCT‐ACC‐CGC‐G and the reverse primer GCG‐CAT‐GAT‐ATT‐CCC‐ATT‐AGC‐AGT‐GC. Reactions consisted of touchdown PCR with denaturation at 95°C for 30 s, annealing at appropriate temperatures for 30 s, and extension at 72°C for 30 s. Finally, the amplified fragments were analyzed by pyrosequencing on a Pyromark Q24 system (QIAGEN) using primer ACC‐CGC‐GGC‐AAG‐ACG.\",\n \"Concordance between Smart Gene assay and conventional diagnosis was evaluated by Cohen's kappa coefficient. All statistical analyses were conducted using the R 4.2.1 software program (www.r‐project.org).\",\n \"[SUBTITLE] Evaluation of “Helicobacter pylori molecular POCT kit” by Smart Gene™ in Helicobacter pylori infection diagnosis [SUBSECTION] Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\\nResults from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\\n[SUBTITLE] Discrepancy in the diagnosis of Helicobacter pylori by Smart Gene [SUBSECTION] The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\\nList of the mismatch cases between Smart Gene™ assay and control methods\\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\\nThe mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\\nList of the mismatch cases between Smart Gene™ assay and control methods\\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\\n[SUBTITLE] Comparison of CAM resistance and the mutation ratio of Helicobacter pylori\\n23S rRNA by pyrosequencing [SUBSECTION] Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\\nQuantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\\n[SUBTITLE] Evaluation of the detection of clarithromycin resistance using “Helicobacter pylori molecular POCT kit” by Smart Gene™\\n [SUBSECTION] “Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\\nComparison of CAM resistance determined by Smart Gene™ assay and AST\\nAbbreviation: AST, Antibiotic susceptibility test.\\n“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\\nComparison of CAM resistance determined by Smart Gene™ assay and AST\\nAbbreviation: AST, Antibiotic susceptibility test.\",\n \"Results from the use of the “H. pylori molecular POCT kit” by Smart Gene™ was strongly correlated with the results from conventional H. pylori diagnosing methods. To evaluate the diagnostic accuracy of “the molecular PCOT kit,” conventional H. pylori diagnostic methods, such as UBT, SAT, culture test, and PCR, were compared (Table 1). The culture test, with a sensitivity of 100% (67/67) and specificity of 95.9% (71/74), showed the highest concordance rate with POCT kit.\\nComparison of Smart Gene™ assay using intragastric fluid and conventional diagnosis of Helicobacter pylori infection\\nAbbreviations: PCR, polymerase chain reaction; SAT, stool antigen test; UBT, urea breath test.\",\n \"The mismatch of “H. pylori molecular POCT kit” by Smart Gene™ from conventional H. pylori diagnosis methods occurred in 19 cases in this study. (Table 2). In most cases, the results of POCT kit did not match completely. However, in several test results, except Case 6, POCT kit gave different results. In Case 6, only the results of POCT kit showed positive. Three cases (Case 4, 5, 13) tested positive for only PCR but negative for other methods including POCT kit.\\nList of the mismatch cases between Smart Gene™ assay and control methods\\nAbbreviations: N.D., no data; PCR, polymerase chain reaction; SAT, Stool antigen test; UBT, UREA BREATH test.\",\n \"Quantitative pyrosequencing analysis of mutation rates at positions 2142 and 2143 of the H. pylori 23S rRNA gene supported the results of “H. pylori molecular POCT kit” by Smart Gene™ in detecting clarithromycin resistance up to a mutation rate of about 20%. The results of POCT kit were compared with quantitative pyrosequencing analysis of the mutation rates of H. pylori 23S rRNA genes at positions 2142 and 2143. Regardless of their locations, all the mutations at position 2142 detected by pyrosequencing analysis were also detected as mutations by POCT kit. On the contrary, although mutations were detected by pyrosequencing analysis at position 2143, POCT kit could only detect the mutations at position 2143 if they were 20% or more (Figure 3). For the mutation at position 2143, the minimum mutation rate determined as a mutation by POCT kit was 15%, while the maximum mutation rate determined as a wild type by POCT kit was 23%.\\nComparison of CAM resistance determined by Smart Gene™ assay and mutation ratio of Helicobacter pylori 23S rRNA gene by pyrosequencing. CAM, clarithromycin\",\n \"“Helicobacter pylori molecular POCT kit” by Smart Gene™ was able to detect clarithromycin resistance up to the mutation rate of about 20% by pyrosequencing. In determining clarithromycin resistance, the concordance rate for resistance was 91.7% (22/24) and the concordance rate for susceptibility was 100% (43/43) between Smart Gene™ and the control test, antibiotic susceptibility test (AST). The overall match rate was 97.0% (65/67; Table 3). Two discrepancy results represented 13% and 23% in the mutation rate through quantitative pyrosequencing analysis, which are low mutation rates.\\nComparison of CAM resistance determined by Smart Gene™ assay and AST\\nAbbreviation: AST, Antibiotic susceptibility test.\",\n \"“Helicobacter pylori molecular POCT kit” by Smart Gene™ allows physicians to safely collect specimens and obtain diagnostic results on time. The simplicity of the kit allows for fast and accurate diagnosis of CAM resistance in determining a patient's precise eradication therapy. Consequently, its wider application will lead to more patients being tested for their H. pylori infection status.\\nThe effectiveness of the “H. pylori molecular POCT kit” in detecting a CAM resistance‐associated mutation rate at a high concordance rate as an antibiotic susceptibility test proves beneficial in daily clinical practice. This novel kit can detect gene mutation at positions 2142 and 2143 of the 23S rRNA gene domain V region of H. pylori, precisely indicating CAM resistance. Furthermore, the “H. pylori molecular POCT kit” can detect CAM resistance‐associated mutations at a mutation rate of 20% or more when compared to pyrosequencing analysis. Results obtained by “H. pylori molecular POCT kit” in detecting CAM resistance is a great indication of patients' eradication results.\\nUsing “H. pylori molecular POCT kit” for detecting CAM‐resistant mutation can significantly improve time‐efficiency in planning a specific treatment plan for patients. Due to its simplicity and speed, compared with other existing antibiotic susceptibility tests, physicians can speed up the process of selecting the appropriate antibiotic for the eradication of H. pylori, allowing patients to receive treatment as early as possible. As a result of the quick results provided by “H. pylori molecular POCT kit,” physicians can provide individualized treatment for patients with CAM resistance, such as skipping CAM‐based therapy and jumping into non‐CAM‐based therapy. Consequently, this will save physicians a significant amount of time they would have wasted on waiting for results and trying out ineffective treatments.\\nAdditionally, “H. pylori molecular POCT kit” is safer for patients than conventional tests because it requires no additional invasive procedure during endoscopic examination and offers painless procedures. Patients only need to provide their intragastric fluid as a test specimen and do not have to go through gastric mucosal biopsy, which could cause complications, including excessive bleeding. Furthermore, because collecting intragastric fluid requires very little preparation time, patients do not have to suffer any pain, which they are likely to experience with longer preparation time. This significantly lower the risk of patient's vital transition.\\nThe ease‐of‐use and reliability of results with the “H. pylori molecular POCT kit” could lead to the consolidation of all the different H. pylori diagnostic tests currently in use. Although the occurrence of false positives and negatives need to be further investigated, the kit has a high likelihood of replacing existing diagnostic tests. In this study, three cases (Case 4, 5, 13) tested positive for only PCR assumed to be the result of insufficient H. pylori amount.\\n“H. pylori molecular POCT kit” is so simple and so capable of accurately and safely diagnosing H. pylori infection that it could be a great alternative diagnostic method in near future. Because the kit has a wide application and allows for the simultaneous testing of H. pylori infection and CAM resistance statuses, physicians can save time on eradication failure caused by CAM resistance.\\nOne limitation of our study is the sample size. Although our sample size was sufficient enough to give us acceptable data, to increase the accuracy of the kit, we would like to enroll and study more people diagnosed with H. pylori infection. Because participating the study and eradicating H. pylori only lower participants' risk of developing gastric cancer, this study is also a part of eradicating overall gastric cancer itself.\",\n \"MT performed the experiments and analyses and drafted the manuscript. YW performed the experiments. RO performed pyrosequencing. RW, MH, and KK performed the experiments. HY and FI supervised the experiments. MK supervised the entire project.\",\n \"MIZUHO MEDY Co., Ltd provided grants and H. pylori molecular POCT kit for this study, but played no role in the study design, data collection and interpretation, or in the decision to submit the work for publication.\",\n \"The authors declare no conflict of interests associated with this study.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"materials-and-methods",null,null,null,null,null,null,null,"results",null,null,null,null,"discussion",null,null,"COI-statement"],"string":"[\n null,\n \"materials-and-methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n \"COI-statement\"\n]"},"keywords":{"kind":"list like","value":["gastric juice","\nH. pylori diagnosis","PCR","point‐of‐care testing","susceptibility test"],"string":"[\n \"gastric juice\",\n \"\\nH. pylori diagnosis\",\n \"PCR\",\n \"point‐of‐care testing\",\n \"susceptibility test\"\n]"}}},{"rowIdx":359,"cells":{"title":{"kind":"string","value":"Proprotein Convertase Subtilisin/Kexin 9 (PCSK9) Promotes Macrophage Activation via LDL Receptor-Independent Mechanisms."},"pmid":{"kind":"string","value":"36263780"},"background_abstract":{"kind":"string","value":"Activated macrophages contribute to the pathogenesis of vascular disease. Vein graft failure is a major clinical problem with limited therapeutic options. PCSK9 (proprotein convertase subtilisin/kexin 9) increases low-density lipoprotein (LDL)-cholesterol levels via LDL receptor (LDLR) degradation. The role of PCSK9 in macrophage activation and vein graft failure is largely unknown, especially through LDLR-independent mechanisms. This study aimed to explore a novel mechanism of macrophage activation and vein graft disease induced by circulating PCSK9 in an LDLR-independent fashion."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We used Ldlr-/- mice to examine the LDLR-independent roles of circulating PCSK9 in experimental vein grafts. Adeno-associated virus (AAV) vector encoding a gain-of-function mutant of PCSK9 (rAAV8/D377Y-mPCSK9) induced hepatic PCSK9 overproduction. To explore novel inflammatory targets of PCSK9, we used systems biology in Ldlr-/- mouse macrophages."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"In Ldlr-/- mice, AAV-PCSK9 increased circulating PCSK9, but did not change serum cholesterol and triglyceride levels. AAV-PCSK9 promoted vein graft lesion development when compared with control AAV. In vivo molecular imaging revealed that AAV-PCSK9 increased macrophage accumulation and matrix metalloproteinase activity associated with decreased fibrillar collagen, a molecular determinant of atherosclerotic plaque stability. AAV-PCSK9 induced mRNA expression of the pro-inflammatory mediators IL-1β (interleukin-1 beta), TNFα (tumor necrosis factor alpha), and MCP-1 (monocyte chemoattractant protein-1) in peritoneal macrophages underpinned by an in vitro analysis of Ldlr-/- mouse macrophages stimulated with endotoxin-free recombinant PCSK9. A combination of unbiased global transcriptomics and new network-based hyperedge entanglement prediction analysis identified the NF-κB (nuclear factor-kappa B) signaling molecules, lectin-like oxidized LOX-1 (LDL receptor-1), and SDC4 (syndecan-4) as potential PCSK9 targets mediating pro-inflammatory responses in macrophages."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Circulating PCSK9 induces macrophage activation and vein graft lesion development via LDLR-independent mechanisms. PCSK9 may be a potential target for pharmacologic treatment for this unmet medical need."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Animals","Mice","Cholesterol","Lipoproteins, LDL","Macrophage Activation","NF-kappa B","Proprotein Convertase 9","Receptors, LDL","Serine Endopeptidases","Subtilisins"],"string":"[\n \"Animals\",\n \"Mice\",\n \"Cholesterol\",\n \"Lipoproteins, LDL\",\n \"Macrophage Activation\",\n \"NF-kappa B\",\n \"Proprotein Convertase 9\",\n \"Receptors, LDL\",\n \"Serine Endopeptidases\",\n \"Subtilisins\"\n]"},"pmcid":{"kind":"string","value":"9973449"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"Animal Procedures"},"methods_text":{"kind":"string","value":"All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner [SUBSECTION] In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\nIn vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\n[SUBTITLE] AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions [SUBSECTION] Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\nEvidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\n[SUBTITLE] AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo [SUBSECTION] We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\nWe previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\n[SUBTITLE] Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\n\n-/- Mice In Vitro [SUBSECTION] To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\nTo support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\n[SUBTITLE] A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages [SUBSECTION] We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\nWe explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\n[SUBTITLE] SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages [SUBSECTION] We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\nWe further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\n[SUBTITLE] Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions [SUBSECTION] Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\nOur study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\n[SUBTITLE] Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development [SUBSECTION] Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\nAlthough our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["What Is Known?","What New Information Does This Article Contribute?","Methods","Culture and Stimulation of Mouse Primary Macrophages","Transcriptomics","Network-Based Hyperedge Entanglement Prediction Analysis","Statistics","AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner","AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions","AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo","Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\n\n-/- Mice In Vitro","A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages","SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages","Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions","Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development","Article Information","Acknowledgments","Sources of Funding","Supplemental Materials"],"string":"[\n \"What Is Known?\",\n \"What New Information Does This Article Contribute?\",\n \"Methods\",\n \"Culture and Stimulation of Mouse Primary Macrophages\",\n \"Transcriptomics\",\n \"Network-Based Hyperedge Entanglement Prediction Analysis\",\n \"Statistics\",\n \"AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner\",\n \"AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions\",\n \"AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo\",\n \"Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\\n\\n-/- Mice In Vitro\",\n \"A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages\",\n \"SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages\",\n \"Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions\",\n \"Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development\",\n \"Article Information\",\n \"Acknowledgments\",\n \"Sources of Funding\",\n \"Supplemental Materials\"\n]"},"other_sections_texts":{"kind":"list like","value":["Autologous saphenous vein is a widely used surgical bypass for lower extremity peripheral artery disease and coronary artery disease.\nPCSK9 (proprotein convertase subtilisin/kexin 9) targets receptors other than LDL receptor (LDLR) such as LRP1 (low-density lipoprotein receptor-related protein-1), ApoER2 (apolipoprotein E receptor 2), VLDLR (very low-density lipoprotein receptor), and CD36.\nPCSK9 may induce pro-inflammatory responses in macrophages and arterial atherosclerotic lesions primarily via LDLR-dependent mechanisms.","PCSK9 promotes vein graft lesion development by the mechanisms unassociated with either LDLR degradation or blood cholesterol levels.\nPCSK9 induces pro-inflammatory macrophage activation: immune responses, proliferation and migration, via LDLR-independent mechanisms.\nPotential PCSK9 targets mediating pro-inflammatory responses in macrophages include NF-κB (nuclear factor-kappa B) signaling molecules, lectin-like oxidized LOX-1 (LDL receptor-1), and SDC4 (syndecan-4).\nThe role of PCSK9 on vascular inflammation remains controversial in the clinical setting because clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity-CRP (C-reactive protein) levels in patients with coronary artery disease. However, CRP may not solely capture changes in inflammation. Although there are a few emerging in vivo studies suggesting a direct link between PCSK9 and atherosclerosis primarily via LDLR-dependent mechanisms, the role of PCSK9 in vein graft failure, especially via LDLR-independent mechanisms, remains largely unknown. To test the hypothesis, we developed a systems approach, involving unbiased transcriptomics and network analysis. Our in vivo evidence suggests that PCSK9 promotes vein graft lesion development, macrophage proliferation, monocyte migration, foam cell formation, and collagen remodeling. We further identified NF-κB signaling molecules, LOX-1 and SDC4 as potential targets of PCSK9 mediating pro-inflammatory responses in macrophages. The evidence, including our own, has suggested that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions.","[SUBTITLE] Data Availability [SUBSECTION] The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\nThe online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\n[SUBTITLE] Animal Procedures [SUBSECTION] All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\nAll animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\n[SUBTITLE] Culture and Stimulation of Mouse Primary Macrophages [SUBSECTION] Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\nMouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\n[SUBTITLE] Transcriptomics [SUBSECTION] RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\nRNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\n[SUBTITLE] Network-Based Hyperedge Entanglement Prediction Analysis [SUBSECTION] Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\nNetwork-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\n[SUBTITLE] Statistics [SUBSECTION] Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\nStatistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.","Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).","RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.","Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.","Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.","In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).","Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.","We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.","To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).","We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.","We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.","Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.","Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.","[SUBTITLE] Acknowledgments [SUBSECTION] We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\nWe thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\n[SUBTITLE] Sources of Funding [SUBSECTION] This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\nThis work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\n[SUBTITLE] Disclosures [SUBSECTION] Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\nPfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\n[SUBTITLE] Supplemental Materials [SUBSECTION] Online Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75\nOnline Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75","We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.","This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.","Online Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75"],"string":"[\n \"Autologous saphenous vein is a widely used surgical bypass for lower extremity peripheral artery disease and coronary artery disease.\\nPCSK9 (proprotein convertase subtilisin/kexin 9) targets receptors other than LDL receptor (LDLR) such as LRP1 (low-density lipoprotein receptor-related protein-1), ApoER2 (apolipoprotein E receptor 2), VLDLR (very low-density lipoprotein receptor), and CD36.\\nPCSK9 may induce pro-inflammatory responses in macrophages and arterial atherosclerotic lesions primarily via LDLR-dependent mechanisms.\",\n \"PCSK9 promotes vein graft lesion development by the mechanisms unassociated with either LDLR degradation or blood cholesterol levels.\\nPCSK9 induces pro-inflammatory macrophage activation: immune responses, proliferation and migration, via LDLR-independent mechanisms.\\nPotential PCSK9 targets mediating pro-inflammatory responses in macrophages include NF-κB (nuclear factor-kappa B) signaling molecules, lectin-like oxidized LOX-1 (LDL receptor-1), and SDC4 (syndecan-4).\\nThe role of PCSK9 on vascular inflammation remains controversial in the clinical setting because clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity-CRP (C-reactive protein) levels in patients with coronary artery disease. However, CRP may not solely capture changes in inflammation. Although there are a few emerging in vivo studies suggesting a direct link between PCSK9 and atherosclerosis primarily via LDLR-dependent mechanisms, the role of PCSK9 in vein graft failure, especially via LDLR-independent mechanisms, remains largely unknown. To test the hypothesis, we developed a systems approach, involving unbiased transcriptomics and network analysis. Our in vivo evidence suggests that PCSK9 promotes vein graft lesion development, macrophage proliferation, monocyte migration, foam cell formation, and collagen remodeling. We further identified NF-κB signaling molecules, LOX-1 and SDC4 as potential targets of PCSK9 mediating pro-inflammatory responses in macrophages. The evidence, including our own, has suggested that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions.\",\n \"[SUBTITLE] Data Availability [SUBSECTION] The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\\nThe online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\\n[SUBTITLE] Animal Procedures [SUBSECTION] All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\\nAll animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\\n[SUBTITLE] Culture and Stimulation of Mouse Primary Macrophages [SUBSECTION] Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\\nMouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\\n[SUBTITLE] Transcriptomics [SUBSECTION] RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\\nRNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\\n[SUBTITLE] Network-Based Hyperedge Entanglement Prediction Analysis [SUBSECTION] Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\\nNetwork-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\\n[SUBTITLE] Statistics [SUBSECTION] Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\\nStatistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\",\n \"Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\",\n \"RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\",\n \"Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\",\n \"Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\",\n \"In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\",\n \"Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\",\n \"We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\",\n \"To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\",\n \"We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\",\n \"We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\",\n \"Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\",\n \"Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\",\n \"[SUBTITLE] Acknowledgments [SUBSECTION] We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\\nWe thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\\n[SUBTITLE] Sources of Funding [SUBSECTION] This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\\nThis work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\\n[SUBTITLE] Disclosures [SUBSECTION] Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\\nPfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\\n[SUBTITLE] Supplemental Materials [SUBSECTION] Online Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\\nOnline Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\",\n \"We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\",\n \"This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\",\n \"Online Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,"methods",null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["What Is Known?","What New Information Does This Article Contribute?","Methods","Data Availability","Animal Procedures","Culture and Stimulation of Mouse Primary Macrophages","Transcriptomics","Network-Based Hyperedge Entanglement Prediction Analysis","Statistics","Results","AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner","AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions","AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo","Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\n\n-/- Mice In Vitro","A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages","SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages","Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions","Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development","Discussion","Article Information","Acknowledgments","Sources of Funding","Disclosures","Supplemental Materials","Supplementary Material"],"string":"[\n \"What Is Known?\",\n \"What New Information Does This Article Contribute?\",\n \"Methods\",\n \"Data Availability\",\n \"Animal Procedures\",\n \"Culture and Stimulation of Mouse Primary Macrophages\",\n \"Transcriptomics\",\n \"Network-Based Hyperedge Entanglement Prediction Analysis\",\n \"Statistics\",\n \"Results\",\n \"AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner\",\n \"AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions\",\n \"AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo\",\n \"Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\\n\\n-/- Mice In Vitro\",\n \"A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages\",\n \"SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages\",\n \"Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions\",\n \"Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development\",\n \"Discussion\",\n \"Article Information\",\n \"Acknowledgments\",\n \"Sources of Funding\",\n \"Disclosures\",\n \"Supplemental Materials\",\n \"Supplementary Material\"\n]"},"all_sections_texts":{"kind":"list like","value":["Autologous saphenous vein is a widely used surgical bypass for lower extremity peripheral artery disease and coronary artery disease.\nPCSK9 (proprotein convertase subtilisin/kexin 9) targets receptors other than LDL receptor (LDLR) such as LRP1 (low-density lipoprotein receptor-related protein-1), ApoER2 (apolipoprotein E receptor 2), VLDLR (very low-density lipoprotein receptor), and CD36.\nPCSK9 may induce pro-inflammatory responses in macrophages and arterial atherosclerotic lesions primarily via LDLR-dependent mechanisms.","PCSK9 promotes vein graft lesion development by the mechanisms unassociated with either LDLR degradation or blood cholesterol levels.\nPCSK9 induces pro-inflammatory macrophage activation: immune responses, proliferation and migration, via LDLR-independent mechanisms.\nPotential PCSK9 targets mediating pro-inflammatory responses in macrophages include NF-κB (nuclear factor-kappa B) signaling molecules, lectin-like oxidized LOX-1 (LDL receptor-1), and SDC4 (syndecan-4).\nThe role of PCSK9 on vascular inflammation remains controversial in the clinical setting because clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity-CRP (C-reactive protein) levels in patients with coronary artery disease. However, CRP may not solely capture changes in inflammation. Although there are a few emerging in vivo studies suggesting a direct link between PCSK9 and atherosclerosis primarily via LDLR-dependent mechanisms, the role of PCSK9 in vein graft failure, especially via LDLR-independent mechanisms, remains largely unknown. To test the hypothesis, we developed a systems approach, involving unbiased transcriptomics and network analysis. Our in vivo evidence suggests that PCSK9 promotes vein graft lesion development, macrophage proliferation, monocyte migration, foam cell formation, and collagen remodeling. We further identified NF-κB signaling molecules, LOX-1 and SDC4 as potential targets of PCSK9 mediating pro-inflammatory responses in macrophages. The evidence, including our own, has suggested that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions.","[SUBTITLE] Data Availability [SUBSECTION] The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\nThe online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\n[SUBTITLE] Animal Procedures [SUBSECTION] All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\nAll animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\n[SUBTITLE] Culture and Stimulation of Mouse Primary Macrophages [SUBSECTION] Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\nMouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\n[SUBTITLE] Transcriptomics [SUBSECTION] RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\nRNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\n[SUBTITLE] Network-Based Hyperedge Entanglement Prediction Analysis [SUBSECTION] Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\nNetwork-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\n[SUBTITLE] Statistics [SUBSECTION] Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\nStatistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.","The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.","All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.","Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).","RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.","Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.","Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.","[SUBTITLE] AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner [SUBSECTION] In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\nIn vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\n[SUBTITLE] AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions [SUBSECTION] Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\nEvidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\n[SUBTITLE] AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo [SUBSECTION] We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\nWe previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\n[SUBTITLE] Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\n\n-/- Mice In Vitro [SUBSECTION] To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\nTo support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\n[SUBTITLE] A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages [SUBSECTION] We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\nWe explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\n[SUBTITLE] SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages [SUBSECTION] We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\nWe further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\n[SUBTITLE] Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions [SUBSECTION] Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\nOur study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\n[SUBTITLE] Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development [SUBSECTION] Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\nAlthough our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.","In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).","Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.","We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.","To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).","We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.","We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.","Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.","Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.","This study investigated LDLR-independent mechanisms by which PCSK9 induces pro-inflammatory activation of macrophages and accelerates vein graft lesion development. The key findings include: (1) AAV-PCSK9 did not change lipid profile but promoted vein graft lesion development in Ldlr-/- mice, suggesting LDLR-independent mechanisms; (2) AAV-PCSK9 increased macrophage accumulation and decreased fibrillar collagen in vein graft lesions; (3) PCSK9 induced macrophage activation in Ldlr-/- mice; (4) unbiased global transcriptomics of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9 revealed activation of several pro-inflammatory pathways; (5) a combination of the global transcriptomics and novel network-based hyperedge entanglement prediction analysis showed NF-κB signaling pathway and OLR1(LOX-1) could play a key role in vein graft lesion development; (6) SDC4, a heparan sulfate proteoglycan, could be a novel target of PCSK9 that mediates pro-inflammatory responses in macrophages.\nReversed autologous saphenous veins are widely used for surgical bypass for treatment of lower extremity peripheral artery disease and coronary artery disease. Peripheral artery disease affects approximately 8.5 million people in the United States aged ≥40 years52 and a total of 202 million worldwide, as of 2010.53 The incidence of peripheral artery disease has been rising by 23.5% globally from 2000 to 2010 largely due to aging populations.53 Of note, 40% of vein grafts for peripheral artery disease are failing or fail within a year.1 The volume of CABG procedures for severe coronary artery disease has declined since 1998; however, 371 000 bypass procedures were still performed in the United States in 2014,52 and approximately 30% of these costly procedures are failing or fail 12 to 18 months after surgery.2\nEvidence suggests that pro-inflammatory responses to PCSK9 in macrophages and arterial atherosclerotic lesions may primarily depend on LDLR. Recombinant PCSK9 induced TNFα mRNA in bone marrow-derived macrophages mainly in an LDLR-dependent fashion.19 Furthermore, overexpression of human PCSK9 in macrophages promoted atherosclerotic lesions.28 Deletion of the PCSK9 gene in the liver reduced atherosclerotic lesions, primarily via mechanisms dependent on LDLR.29 To the best of our knowledge, no in vivo studies have demonstrated LDLR-independent pro-inflammatory effects of PCSK9 on vascular lesion development. The present study demonstrated that circulating PCSK9 induces macrophage activation and vein graft lesion development in LDLR-independent mechanisms. The relative contributions of the LDLR-dependent and independent mechanisms, however, remain unknown.\nClinical reports demonstrated that cholesterol-lowering with PCSK9 inhibitors reverses the pro-inflammatory profile of monocytes in patients with hypercholesterolemia,54 indicating the lipid-dependent pro-inflammatory effects of PCSK9 in humans. Clinical trials also demonstrated that PCSK9 inhibitors lower not only serum levels of LDL cholesterol but also lipoprotein (a) (Lp(a)).55 Mice lack apolipoprotein (a) or Lp(a),56 which enabled us to rule out the role of Lp(a) in the effects of PCSK9 in our in vivo study. Although genetic evidence suggests LDL-dependent effects of PCSK9 in gain-of-function and loss-of-function carriers57 and the clinical benefits of antiPCSK9 strategies derive largely from LDL lowering, nonLDL-dependent effects might also contribute.58 As serum total cholesterol and triglyceride levels did not differ in AAV-LacZ and AAV-PCSK9 groups, our findings in Ldlr-/- mice do not seem to depend on PCSK9’s effects on the profile of circulating lipids.\nHigh sensitivity C-reactive protein is the most established marker of inflammation to predict cardiovascular events independently in primary and secondary prevention.59 Clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity C-reactive protein levels in patients with coronary artery disease.60,61 Antiinflammatory changes in monocyte phenotype with PCSK9 antibodies were not accompanied by a CRP (C-reactive protein) reduction in patients with familial hypercholesterolemia, suggesting that CRP may not solely capture changes in local inflammation produced by PCSK9 inhibition.54 In addition, the ATHEROREMO-IVUS study (The European Collaborative Project on Inflammation and Vascular Wall Remodeling in Atherosclerosis - Intravascular Ultrasound) demonstrated a linear relationship between serum levels of PCSK9 and the fraction or amount of the necrotic core in the nonculprit lesions of patients with acute coronary syndrome, independently of serum levels of LDL cholesterol. Our study provides support for the notion that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions. Further studies are needed to explore whether PCSK9 inhibition ameliorates local inflammation in vascular lesions in patients.\nIn general, multi-omics technologies, such as proteomics and transcriptomics, identify many differentially expressed molecules. It is challenging to assess direct interactions between these molecules or their causal roles in a biological or disease context by examining each molecule or interaction among many. This challenge has driven efforts in computational network biomedicine to develop unbiased prediction methods. By exploiting the mere topology of a protein-protein interaction network, one can quantify the likelihood of direct interactions between a molecule of interest (eg, PCSK9 in the present study) and a hyperedge of differentially expressed molecules (which represents a network functional module). We used a novel node2hyperedge entanglement method20 and showed any of the 24 differentially expressed transcripts, including NF-κB signaling molecules, OLR1(LOX-1), and SDC4, as a potential target of PCSK9 associated with inflammation. In the present study, we identified SDC4 as a potential novel target of PCSK9. Although Sdc4 was not a high-ranking differential expressed transcript, its fold change after stimulation with recombinant PCSK9 in the RNA-sequencing was statistically significant. We further performed in vitro experiments involving siRNA silencing in primary macrophages to substantiate our computational prediction platform and demonstrated that SDC4 indeed binds to PCSK9 in vivo and mediates PCSK9-induced pro-inflammatory responses. The contribution of macrophage SDC4 to PCSK9-induced pro-inflammatory responses in vivo remains to be elucidated. This novel prediction method may offer a powerful and necessary tool for the completion of the human interactome and for generating new hypotheses.\nSDC4 regulates several robust pro-inflammatory molecules, such as IL-1β, NLRP3 (NLR family pyrin domain containing 3), CD40 (cluster of differentiation 40) and ICAM1. Various endogenous signals abundantly present in atherosclerotic lesions activates NLRP3, a key mediator of IL-1β maturation.62,63 T cell-mediated macrophage activation involves CD40 to produce interstitial collagens and tissue-factor pro-coagulant.64 The role of macrophage ICAM1 in atherosclerosis remains obscure, but a recent study reported that macrophage ICAM1 suppresses “M2-like” macrophages during tumor progression. SDC4 silencing inhibited these pro-inflammatory molecules. The magnitude of inhibition, however, was modest, suggesting that while we propose SDC4 is a novel target of PCSK9, but PCSK9 may also target other molecules (eg, NF-κB signaling molecules). The mechanisms mediated via SDC4 we report here appears to be one of multiple pathways affected by PCSK9. Heparan sulfate proteoglycans modulate macrophage activation by maintaining cells in a quiescent state through the capture and sequestration of IFNβ.65,66 We speculate that SDC4 serves as a mediator of PCSK9-induced pro-inflammatory responses through the capture of PCSK9 on the surface of macrophages as seen in the hepatocyte.43 Our data demonstrated that NF-κB may mediate pro-inflammatory responses downstream of the PCSK9 and SDC4 binding. Understanding more detailed mechanisms by which SDC4 mediates the pro-inflammatory responses and their downstream signaling induced by PCSK9 requires future investigations.\nOur study does not demonstrate definitively that local PCSK9 plays an important role in vein graft lesion development. In situ hybridization and qPCR data suggest that SMCs may be the primary source of PCSK9 in these lesions. The expression level of PCSK9 in SMCs was much lower than in the liver, although the relative gene expression level may not be a valid determinant of its biological role or importance. Whether and how circulating PCSK9 affects lesional macrophages through the contact with endothelial cells (which in turn affects macrophage phenotype or the entry into the subendothelial space, which results in direct contact with macrophages) remains to be elucidated.67 Future work needs to define detailed molecular mechanisms.\nOur data do not demonstrate that human primary macrophages can express PCSK9, even when stimulated with pro-inflammatory stimuli, which was consistent with previous studies.45,68 Evidence suggests, however, the participation of endogenous PCSK9 in macrophages during pro-inflammatory responses. In vitro experiments showed that overexpression of PCSK9 in macrophage-like THP-1 cells enhanced the pro-inflammatory response induced by oxidized LDL via NF-κB activation.38 In contrast, PCSK9 gene silencing inhibited the response.38,47 In vivo evidence also showed that overexpression of human PCSK9 in macrophages increased atherosclerotic lesion formation in Apoe-/- mice.28 These results suggest that not only circulating PCSK9 produced from the liver but also local PCSK9 produced from SMCs and macrophages may induce a pro-inflammatory response in macrophages and contribute to vascular lesion development. What cell types other than SMCs contribute to monocyte/macrophage infiltration, and activation remains to be elucidated.\nThe present study demonstrates several lines of novel evidence that circulating PCSK9 promotes macrophage activation and vein graft lesion development via mechanisms independent of blood cholesterol and LDLR degradation. Despite current therapeutics and research advancements, effective medical solutions are limited to prevent vein graft failure. Our study provides molecular bases that circulating PCSK9 deserves investigations as a potential therapeutic target for vein graft failure.","[SUBTITLE] Acknowledgments [SUBSECTION] We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\nWe thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\n[SUBTITLE] Sources of Funding [SUBSECTION] This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\nThis work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\n[SUBTITLE] Disclosures [SUBSECTION] Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\nPfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\n[SUBTITLE] Supplemental Materials [SUBSECTION] Online Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75\nOnline Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75","We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.","This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.","Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.","Online Supplemental Materials and Methods\nSupplemental Figures 1–14\nSupplemental Tables 1–4\nReferences 69–75",""],"string":"[\n \"Autologous saphenous vein is a widely used surgical bypass for lower extremity peripheral artery disease and coronary artery disease.\\nPCSK9 (proprotein convertase subtilisin/kexin 9) targets receptors other than LDL receptor (LDLR) such as LRP1 (low-density lipoprotein receptor-related protein-1), ApoER2 (apolipoprotein E receptor 2), VLDLR (very low-density lipoprotein receptor), and CD36.\\nPCSK9 may induce pro-inflammatory responses in macrophages and arterial atherosclerotic lesions primarily via LDLR-dependent mechanisms.\",\n \"PCSK9 promotes vein graft lesion development by the mechanisms unassociated with either LDLR degradation or blood cholesterol levels.\\nPCSK9 induces pro-inflammatory macrophage activation: immune responses, proliferation and migration, via LDLR-independent mechanisms.\\nPotential PCSK9 targets mediating pro-inflammatory responses in macrophages include NF-κB (nuclear factor-kappa B) signaling molecules, lectin-like oxidized LOX-1 (LDL receptor-1), and SDC4 (syndecan-4).\\nThe role of PCSK9 on vascular inflammation remains controversial in the clinical setting because clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity-CRP (C-reactive protein) levels in patients with coronary artery disease. However, CRP may not solely capture changes in inflammation. Although there are a few emerging in vivo studies suggesting a direct link between PCSK9 and atherosclerosis primarily via LDLR-dependent mechanisms, the role of PCSK9 in vein graft failure, especially via LDLR-independent mechanisms, remains largely unknown. To test the hypothesis, we developed a systems approach, involving unbiased transcriptomics and network analysis. Our in vivo evidence suggests that PCSK9 promotes vein graft lesion development, macrophage proliferation, monocyte migration, foam cell formation, and collagen remodeling. We further identified NF-κB signaling molecules, LOX-1 and SDC4 as potential targets of PCSK9 mediating pro-inflammatory responses in macrophages. The evidence, including our own, has suggested that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions.\",\n \"[SUBTITLE] Data Availability [SUBSECTION] The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\\nThe online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\\n[SUBTITLE] Animal Procedures [SUBSECTION] All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\\nAll animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\\n[SUBTITLE] Culture and Stimulation of Mouse Primary Macrophages [SUBSECTION] Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\\nMouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\\n[SUBTITLE] Transcriptomics [SUBSECTION] RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\\nRNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\\n[SUBTITLE] Network-Based Hyperedge Entanglement Prediction Analysis [SUBSECTION] Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\\nNetwork-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\\n[SUBTITLE] Statistics [SUBSECTION] Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\\nStatistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\",\n \"The online-only Supplemental Data provided detailed methods of all procedures. The data, analytic methods, and study materials will be made available to other researchers, upon request for purposes of reproducing the results or replicating the procedure.\",\n \"All animal experiments were approved by the Brigham and Women’s Hospital’s Animal Welfare Assurance (protocol 2016N000219). Male Ldlr-/- mice were fed a high-fat diet (1.25% cholesterol, D12108C, Research Diets, Inc., New Brunswick, NJ). To create experimental vein grafts in Ldlr-/- mice, inferior vena cava were harvested from a donor mouse and implanted into the carotid artery of recipient mice using cuff technique, as previously described.8,21 To induce liver-specific gain-of-function mutant PCSK9, we administered adeno-associated virus (AAV) encoding a gain-of-function form of mouse PCSK9 (AAV-PCSK9, 1×1011 vg, i.v.).22 We constructed AAV encoding LacZ (AAV-LacZ) that we used as a control for AAV-PCSK9. Four weeks after implantation, grafts were evaluated. After ultrasonography and in vivo molecular imaging of vein grafts, vein graft tissues were harvested, sectioned, and used for Masson trichrome staining, picrosirius red staining, immunohistochemistry, and in situ hybridization.\",\n \"Mouse peritoneal macrophages cultured with RPMI 1640 containing 10% fetal bovine serum were prepared as previously described.8 Briefly, 4% Brewer thioglycolate medium (BD Diagnostic Systems) was injected into the peritoneal cavity of mice 4 days before macrophage collection. After euthanasia, 10 mL of ice-cold phosphate-buffered saline was injected into the peritoneal cavity, and cells were harvested. The cells were then washed with phosphate-buffered saline once and plated for further experiments. Overnight starvation was performed with 0.5% fetal bovine serum before each experiment. Endotoxin-free recombinant mouse PCSK9 derived from mouse myeloma cell line NS0 was used to stimulate mouse primary macrophages (R&D Systems).\",\n \"RNA-sequencing—mRNAseq (polyA enriched) library prep, single-end 75bp sequencing on NextSeq of 12-16 pooled barcoded samples, and VIPER analysis23—was performed at the Molecular Biology Core Facilities at Dana-Farber Cancer Institute. The data were then analyzed using DESeq2 and Qlucore (http://www.qlucore.com/) to perform a 2-group comparison and identify differentially expressed transcripts (increase/decrease by 2-fold) between control and PCSK9 groups. STRING (version 11.5) was used to evaluate known and predicted transcript interconnectivity. The differentially expressed transcripts (increase/decrease by 2-fold, q<0.05), which were compared with the “inflammatome”24 dataset, were used for the following network analysis.\",\n \"Network-Based Hyperedge Entanglement Prediction (HEP) algorithm20 was performed to estimate the likelihood of PCSK9 directly interacting with any of the differentially expressed transcripts in the “inflammatome.” Given only a network topology and its high-order hypergraph organization, the HEP algorithm uses an ensemble of link predictors to calculate a level of statistical significance for each candidate node to hyperedge entanglement that is characterized by not having a direct connection. The link predictor is an operator that, exploiting topological information of the graph, can associate a similarity score to any disconnected node pair, suggesting the likelihood for a link between them to exist. Several link predictors perform under the evidence that 2 nodes are more likely to be linked if their common neighbors are members of a strongly inner-linked cohort, named a local community,25 and recent studies demonstrated that this growth mechanism emerges mainly considering missing links between nonadjacent nodes distanced by paths of length 2 (L2) or 3 (L3)26 following a scheme that can be well described by the Cannistraci-Hebb network automata modeling.27 The HEP analysis provides 12 different estimations of the P value that quantifies the significance of the entanglement (association) between each node and hyperedge (also called node2hyperedge entanglement) using the 12 available node-node link prediction variants.25, 27 In this study, we applied HEP algorithm according to the following specifications: the scaling of the similarity scores was not necessary, and the average operator to estimate the node2hyperedge entanglement scores was the mean.\",\n \"Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). For experiments with a small sample size (n<6), P values were determined by nonparametric analysis except for RNA-sequencing analysis. Other data were examined for normality before analysis by Shapiro-Wilk test. For all normally distributed data, an unpaired 2-tailed Student t test was used for comparisons between 2 groups, 1-way ANOVA followed by Bonferroni or Dunnett post-hoc test was used to compare multiple groups. When there were 2 experimental factors, 2-way ANOVA for comparisons between multiple groups was used. If the data were not normally distributed, nonparametric unpaired 2-tailed Mann-Whitney U test was used to compare 2 groups, and Kruskal-Wallis test with Dunn post-hoc test was used to compare multiple groups. Wald Test in DESeq2 was used for bulk RNA-sequencing analysis. Data are expressed as mean ± SEM for continuous variables. P values of <0.05 were considered statistically significant.\",\n \"[SUBTITLE] AAV-PCSK9 Promotes Vein Graft Lesion Development in an LDLR-Independent Manner [SUBSECTION] In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\\nIn vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\\n[SUBTITLE] AAV-PCSK9 Increases Thin Collagen Fibers and Macrophage Accumulation in Vein Graft Lesions [SUBSECTION] Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\\nEvidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\\n[SUBTITLE] AAV-PCSK9 Induces Macrophage Activation via an LDLR-Independent Route In Vivo [SUBSECTION] We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\\nWe previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\\n[SUBTITLE] Endotoxin-Free Recombinant PCSK9 Induced Macrophage Activation in Peritoneal Macrophages of Ldlr\\n\\n-/- Mice In Vitro [SUBSECTION] To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\\nTo support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\\n[SUBTITLE] A combination of Unbiased Global Transcriptomics and Network-Based Hyperedge Entanglement Prediction Analysis Revealed Potential Targets of PCSK9 in Macrophages [SUBSECTION] We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\\nWe explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\\n[SUBTITLE] SDC4 Silencing Suppresses PCSK9-Induced Pro-Inflammatory Responses in Macrophages [SUBSECTION] We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\\nWe further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\\n[SUBTITLE] Local Production of PCSK9 in the Liver Was Greater Than in Vein Graft Lesions [SUBSECTION] Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\\nOur study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\\n[SUBTITLE] Macrophages Rather Than Endothelial Cells and SMCs May Mediate the Effects of PCSK9 on Vein Graft Lesion Development [SUBSECTION] Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\\nAlthough our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\",\n \"In vivo studies have suggested a direct link between PCSK9 and atherosclerosis28,29; however, no evidence has yet implicated PCSK9 in vein graft lesion development. Furthermore, these in vivo studies indicate that the underlying mechanism of atherosclerotic lesion development by PCSK9 may depend mainly on LDLR. Whether PCSK9 can exert pro-inflammatory effects via LDLR-independent mechanisms is unclear. To test the hypothesis that circulating PCSK9 promotes vein graft lesion development in vivo, we used experimental vein grafts in mice.21 A single intravenous injection of AAV-PCSK9 induced liver production of PCSK9. To exclude potential systemic toxicity of AAV, AAV-LacZ was developed and used as a control (Figure 1A; Online Supplemental Methods). All experiments used Ldlr-/- mice to explore LDLR-independent mechanisms unless otherwise noted.\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on vein graft lesion development in Ldlr-/- mice. A, An experimental protocol for the treatments in Ldlr-/- mice. B, In situ hybridization of PCSK9 (red) in the liver of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9. Nuclei are visualized with DAPI. The data represent 4 mice per group. Scale bars indicate 100 μm. C, PCSK9 mRNA level in the liver of AAV treated Ldlr-/- mice (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively), and serum levels of PCSK9, total cholesterol, and triglyceride in AAV treated Ldlr-/- mice (n=11 per group). D, Left, ultrasonography images of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. The red, green, and yellow dotted lines indicate lumen, near wall, and far wall, respectively. Right, The quantitative analysis of lumen and vessel wall area in short axis view and 3D reconstructed vessel wall volume of vein grafts by ultrasonography (n=12 and 11 for AAV-LacZ and AAV-PCSK9 groups, respectively). E, Left, Masson’s trichrome staining of vein grafts from Ldlr-/- mice treated with AAV-LacZ or AAV-PCSK9 28 days after implantation. Scale bars indicate 500 μm. Right, The quantitative analysis of intimal area, intimal and media/adventitia thickness, and lumen and vessel diameter of vein grafts (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). P value was calculated by Mann-Whitney U-test (triglyceride in C, media/adventitia thickness, lumen diameter, and vessel diameter in E) and unpaired Student t test (A, B, and D). Data are reported as mean ± SEM.\\nBody weight and blood pressure did not differ significantly between the AAV-LacZ and AAV-PCSK9 groups 4 weeks after vein graft implantation (Supplemental Table 1). In situ hybridization showed that AAV-PCSK9 markedly increased PCSK9 mRNA levels in the liver (Figure 1B). AAV-PCSK9 increased PCSK9 mRNA expression in the liver and serum levels of PCSK9 (Figure 1C; Supplemental Table 1). Serum levels of total cholesterol and triglycerides did not differ between the AAV-LacZ and AAV-PCSK9 groups (Figure 1C; Supplemental Table 1). These findings indicate that the effects of AAV-PCSK9 on vein graft lesions described below did not depend on blood cholesterol and triglyceride concentrations (Supplemental Table 1).\\nNoninvasive ultrasonography visualized that AAV-PCSK9 injection caused increased vessel wall area, but not lumen area, in the short axis view and 3-dimensional reconstructed vessel wall volume of vein grafts (Figure 1D). Consistent with the ultrasonography, histological analysis by Masson trichrome staining showed that AAV-PCSK9 increased intimal area and thickness of vein grafts. Lumen diameter, vessel diameter, and media/adventitia thickness did not differ significantly, suggesting compensatory outward remodeling of vein grafts (Figure 1E).\",\n \"Evidence suggests that vein grafts in humans can develop lesions similar to those of advanced arterial atherosclerosis, and plaque rupture can occur in inflamed vein grafts.30,31 Therefore, exploring vein graft lesion composition in depth is critical to understanding what contributes to this occurrence. We then examined the collagen content of vein grafts by picrosirius red staining. AAV-PCSK9 increased the percentage of thin collagen fibers as a ratio of the total collagen fibers in the vein grafts, suggesting that circulating PCSK9 induced pathological features similar to those in thin-capped atherosclerotic plaques (Figure 2A). In vivo molecular imaging further provided insight into the effects of circulating PCSK9 on matrix metalloproteinase activity and macrophage accumulation in vein grafts. We intravenously co-injected 2 spectrally different imaging agents that elaborate near-infrared signals for visualization of matrix metalloproteinase activity (MMPsense, 680 nm) and macrophage accumulation (AminoSPARK, 750 nm). Following the evidence that activated macrophages express MMPs that degrade interstitial collagen as previously reported by our group,32,33 in vivo molecular imaging demonstrated that increased matrix metalloproteinase activity co-localized with accumulated macrophages (Figure 2B, left). In addition, AAV-PCSK9 treatment promoted matrix metalloproteinase activity and macrophage accumulation in vein grafts (Figure 2B), suggesting a potential mechanism of collagen remodeling as determined by Picrosirius red staining. Immunohistochemical analysis further demonstrated increased macrophage accumulation in vein grafts by AAV-PCSK9 (Mac3; Figure 2C). Macrophage proliferation and migration may contribute to the mechanisms of macrophage accumulation in vascular diseases.34 Supporting this mechanism, AAV-PCSK9 increased the percentage of macrophages positive for Ki-67, an indicator of proliferating cells (Supplemental Figure 1A, B). Consistent with this in vivo finding, recombinant PCSK9 augmented M-CSF-induced cell proliferation in mouse bone marrow-derived macrophages (BMDMs) (Supplemental Figure 2A, B). Monocyte migration assay using the human monocytic cell line THP-1 showed that recombinant PCSK9 tended to promote MCP-1 (monocyte chemoattractant protein-1)-induced chemotaxis, suggesting lipid-independent effects of PCSK9 on monocyte migration (Supplemental Figure 2C). On the other hand, AAV-PCSK9 did not increase apoptosis determined by TUNEL staining in vein grafts of Ldlr-/- mice (Supplemental Figure 3A). Accordingly, AAV-PCSK9 did not change necrotic core area in the vein grafts (Supplemental Figure 3B).\\nThe effects of Adeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) on collagen thinning and macrophage accumulation in vein graft lesions of Ldlr-/- mice. A, Picrosirius red staining of vein grafts without (top) or with (bottom) polarized light (n=12 and 11 for AAV-LacZ and AAV-PCSK9 group, respectively). Scale bars indicate 500 μm. Circle graphs show the percentage of thin (green) and thick (red) collagen fibers compared with total fibers. B, Intravital microscopy images of MMPSense 680 (red) and AminoSPARK750 (green) in vein grafts for visualization of MMP activity and macrophage accumulation, respectively (n=10 and 9 for AAV-LacZ and AAV-PCSK9 group, respectively). C, Mac3 (macrophages) staining in vein grafts. Scale bars indicate 500 μm. P value was calculated by Mann-Whitney U-test (thick collagen in A, macrophage accumulation mean fluorescence intensity in B) and unpaired Student t test (C). Data are reported as mean ± SEM.\",\n \"We previously reported that macrophage activation plays a key role in vein graft lesion development.8 To explore the effects of circulating PCSK9 on macrophage activation, we examined mRNA levels of pro-inflammatory (IL-1β [interleukin-1 beta], IL-6, TNFα, and MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1 [transforming growth factor-beta 1], and IL-10) in the peritoneal F4/80+ macrophages of Ldlr-/- mice 1 or 4 weeks after intravenous injection of AAVs. AAV-PCSK9 promoted pro-inflammatory IL-1β, IL-6, TNFα, and MCP-1 mRNA levels at 1 week (Figure 3A). AAV-PCSK9 also suppressed Arginase-1 mRNA levels at 4 weeks when the increase in the mRNA levels of pro-inflammatory molecules at 1 week had subsided (Figure 3B). We have recently reported that predominant pathways of the vein grafts change over time during lesion development. In the same study, pathways with immune responses represent the early responsive proteins in proteomics analysis.9 These results indicate that PCSK9 can induce pro-inflammatory responses in macrophages initially and later impair resolution of inflammation in vivo.\\nAdeno-associated virus (AAV)-PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in murine peritoneal macrophages via LDL receptor (LDLR)-independent mechanisms in vivo. mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages of Ldlr-/- mice 1 week (A) or 4 weeks (B) after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 to 8 per group). P value was calculated by Mann-Whitney U-test (TNFa in A, IL-1b, TNFa, MCP-1, Arginase-1, Ym1, and IL-10 in B) and unpaired Student t test. Data are reported as mean ± SEM.\",\n \"To support our in vivo evidence, we examined in vitro the effects of physiologically relevant levels of recombinant mouse PCSK9 (0.1–2.5 μg/mL)35 on the mRNA levels of the same pro-inflammatory and antiinflammatory molecules examined in Ldlr-/- mouse macrophages. First, we conducted a time-course study of recombinant PCSK9-induced TNFα and Arginase-1 mRNA expression levels to determine an optimal time point for harvesting Ldlr-/- mouse macrophages. The increase of TNFα mRNA levels peaked 3 hours after stimulation, whereas the decrease of Arginase-1 mRNA levels plateaued 12 hours after stimulation with recombinant PCSK9 (Figure 4A). For all experiments, Ldlr-/- mouse macrophages were therefore harvested 3 and 12 hours after stimulation to measure pro- and antiinflammatory molecules, respectively. Peritoneal macrophages treated with recombinant PCSK9 exhibited increased mRNA levels of IL-1β, IL-6, TNFα, and MCP-1 in a concentration-dependent manner (Figure 4B). In contrast, recombinant PCSK9 did not decrease the mRNA levels of antiinflammatory molecules (Figure 4C). These results support the in vivo evidence of macrophage activation by circulating PCSK9 in an LDLR-independent fashion.\\nRecombinant PCSK9 (proprotein convertase subtilisin/kexin 9) induced macrophage activation in LDL receptor (LDLR)-independent mechanisms in vitro. A, Time course of TNFα and Arginase-1 mRNA levels in peritoneal macrophages of wild-type mice after stimulation with recombinant mouse PCSK9 (MmPCSK9; 2.5 μg/mL) (n=5 per group). P value was calculated by Kruskal-Wallis test, followed by Dunn multiple comparison test. B, mRNA levels of pro-inflammatory molecules (IL-1β, IL-6, TNFα, MCP-1) and (C) antiinflammatory molecules (Arginase-1, Ym1, MRC1, TGF-β1, IL-10) were measured in murine peritoneal macrophages from Ldlr-/- mice 3 and 12 hours after stimulation with recombinant mouse PCSK9 for pro-inflammatory and antiinflammatory molecules, respectively (n=9 per group). P value was calculated by 1-way ANOVA, followed by Dunnett multiple comparison test (IL-6, TNFa, MCP-1‚ MRC1) or Kruskal-Wallis test, followed by Dunn multiple comparison test. Data are reported as mean ± SEM.\\nEndotoxin contamination in recombinant proteins can induce pro-inflammatory responses. Therefore, we chose a recombinant protein derived from a mammalian system, mouse myeloma cell line (NS0). The endotoxin level of recombinant mouse PCSK9 used in this study was 0.00441 EU/µg or less, well below the threshold of 0.1 EU/µg required for accurate cell-based assays.19 To exclude further endotoxin contamination in the recombinant PCSK9 protein affecting any measurements, we performed the chromogenic Limulus amebocyte lysate endotoxin assay. The final endotoxin level of the maximum dose of the recombinant protein (2.5 µg/mL) was 0.005 EU/mL or less, which coincides with the endotoxin levels of the commercially available endotoxin-free medium. In addition, mass spectrometry conducted on the solvent derived from reconstituted recombinant mouse PCSK9 protein solution did not produce signals, suggesting that there were no small molecule impurities present (Supplemental Figure 4A). Furthermore, stimulation with heat-inactivated recombinant mouse PCSK9 protein (2.5 μg/mL) caused no changes in the mRNA levels of IL-1β, TNF-α, and MCP-1, demonstrating that endotoxin contamination did not drive the effects observed (Supplemental Figure 4B).\",\n \"We explored further LDLR-independent pro-inflammatory signaling pathways in an unbiased manner, using RNA-sequencing of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. When analyzing the differentially expressed transcripts (FC>2.0, q<0.001), we observed 46 transcripts significantly increased by PCSK9 stimulation in Ldlr-/- macrophages compared with controls (Figure 5A). A hierarchical heat map of Ldlr-/- mouse macrophages filtered by differentially expressed transcripts showed a clear separation between control (no stimulus) and PCSK9 group (Figure 5B). We provided statistically significantly overrepresented transcripts ranked according to their fold change abundance (Supplemental Table 3A) to further examine these differentially expressed transcripts. PCSK9 treatment increased many transcripts implicated in pro-inflammatory responses in Ldlr-/- macrophages (eg, IL-1b, TNF, CXCL2, NF-κB [nuclear factor-kappa B]-related genes).19\\nTranscriptomics, network analysis, and hyperedge entanglement prediction (HEP) of mouse macrophages stimulated with recombinant PCSK9 (proprotein convertase subtilisin/kexin 9). A, Volcano plot for Ldlr-/- mouse macrophages stimulated with recombinant PCSK9. Red markers indicate significantly expressed transcripts with a fold change cutoff of 2.0 and adjusted P value cutoff of 0.001. The enriched transcripts that were previously reported were denoted with their names. B, Heat map of the transcriptomics of Ldlr-/- mouse macrophages (FC>2, q<0.001-filtered data for the difference between macrophages stimulated with and without recombinant PCSK9, n=4 per group). C, The schematic network visualization by HEP prediction. The significantly predicted association of PCSK9 to the 24 mapped inflammation-related transcripts derived from the list of transcripts (FC>2.0, q<0.05) indicates every unknown pairs of interaction (dashed line) that have to be proven yet experimentally. One of them was validated by silencing Sdc4 (syndecan-4) mRNA (blue dashed line). All other links represent known physical interactions (solid gray lines). The node size illustrates the degree of the proteins in the protein-protein interaction network. D, Network analysis of the 24 overrepresented transcripts in Ldlr-/- mouse macrophages using STRING database. Nodes are colored according to k-means clustering (number of clusters=3). Only connected nodes are shown here.\\nTo provide additional mechanistic links, we attempted to identify PCSK9-regulated proteins other than LDLR that mediate the aforementioned pro-inflammatory downstream responses. Many cascading effects in the transcriptional regulatory machinery could control the differentially expressed mRNAs. This set of differentially expressed transcripts were identified as the result of a genome-wide transcriptional profiling method, which might not necessarily interact directly with PCSK9. To estimate the likelihood of PCSK9 directly interacting with those differentially expressed transcripts, we tested this hypothesis verification through network computational analysis. In brief, PCSK9 can be seen as a node in a protein-protein interaction network,36 and the set of differentially expressed transcripts can be seen as a hyperedge projected on the protein-protein interaction network (a group of nodes that are associated by a common feature of biological functions). Muscoloni et al recently introduced a novel algorithm for HEP that exploits an ensemble of link predictors and that provides a level of statistical significance for the entanglement between each node and hyperedge that does not have a direct interaction in the network.20 We then performed the HEP analysis to quantify the significance of the entanglement between PCSK9 and the hyperedge of 24 differentially expressed transcripts (FC>2.0, q<0.05) in the dataset (Supplemental Figure 5A and 5B, Supplemental Table 3B). The resulting P values of 7/12 node-node link prediction methods (RA, CH2-L2, CH3-L2, and all 4 L3-based methods) established a significant association between PCSK9 and the hyperedge of the 24 differentially expressed transcripts (Figure 5C; Supplemental Table 2). These data suggest that any of the 24 differentially expressed transcripts have a high likelihood of directly interacting with PCSK9, enabling us to validate a direct interaction between PCSK9 and one of these 24 differentially expressed transcripts. To understand the biological relevance of these transcripts, we explored the network connectivity of these transcripts identified by HEP analysis using the STRING database (confidence interaction score ≥0.7).37 The network analysis showed a highly clustered network (average local clustering coefficient: 0.547) containing 24 nodes with 52 edges (expected number of edges=3), providing significantly more interactions than expected for a random set of genes of similar size (P value ≤1.0e−16). The k-means clustering (number of clusters=3) showed that one of these 3 clusters mainly consisted of the NF-κB signaling pathway based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (IL-1b, CXCL1 [chemokine (C-X-C motif) ligand 1], CXCL2 [chemokine (C-X-C motif) ligand 2], ICAM1, VCAM1 [vascular cell adhesion molecule 1], RELB [RELB Proto-Oncogene, NF-kB subunit], NFKB2 [nuclear factor kappa B subunit 2]), and the other 2 clusters included OLR1 (LOX-1) and SDC4, respectively (Figure 5D).\\nIt is already reported that NF-κB signaling pathway is involved in PCSK9-induced atherosclerotic inflammation.38 In the present study, AAV-PCSK9 promoted NF-κB p65 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Supplemental Figure 6A). Inhibition of NF-κB signaling pathway with IκB kinase inhibitor, 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA1), decreased pro-inflammatory responses to recombinant PCSK9 in Ldlr-/- mouse macrophages (Supplemental Figure 6B). These data indicate that NF-κB may mediate PCSK9-induced pro-inflammatory responses in vein graft lesions.\\nPCSK9 stimulates the expression of LOX-1, which in turn takes up ox-LDL in macrophages.39 AAV-PCSK9 increased foam cell formation determined by Oil red O positive area in vein grafts (Supplemental Figure 7A). We further found that recombinant PCSK9 increased ox-LDL uptake in bone marrow-derived macrophages from Ldlr-/- mice (Supplemental Figure 7B). These results indicate that PCSK9 also induces foam cell formation by increased ox-LDL uptake, contributing not only to atherosclerotic but also to vein graft lesion development.\\nAmong the members of this differentially expressed transcript hyperedge, we found that SDC4 included in the last cluster could be a novel target of PCSK9 (Figure 5D). While SDC4 was not a high-ranking molecule in the transcript ranking lists (30th in Supplemental Table 3A and 67th in Supplemental Table 3B), its fold change after stimulation with recombinant PCSK9 in Ldlr-/- mice macrophages was statistically significant (2.65, log2FC>1). AAV-PCSK9 also increased SDC4 mRNA level in the peritoneal F4/80+ macrophages of Ldlr-/- mice (Figure 6A). SDC4 is a heparan sulfate proteoglycan40 expressed on the surface of human macrophages.41 In addition, SDC4 mRNA is increased in M(LPS) but not M(IL-4/IL-13) or M(IL-10) in bone marrow-derived macrophages.42 A recent study reported that heparan sulfate proteoglycans physically bind to PCSK9 on the hepatocyte surface43 However, the investigators did not address the interaction between PCSK9 and any specific individual syndecans. Our in silico protein-protein docking analysis using HPEPDOCK, a web server for protein-protein docking based on a hierarchical algorithm, predicted binding between PCSK9 and SDC4 (Supplemental Figure 8A).44 We entered the pdb files of PCSK9 (PDB ID: 2PMW), LDLR (PDB ID: 1N7D), and SDC4 (PDB ID: 1EJP) into HPEPDOCK server, which in turn presented the top 100 docking models based on docking energy minimized scores. Protein-protein docking analysis showed that the binding efficiency of SDC4 and PCSK9 (−212.23 to −235.6 kJ/mol) was comparable to that of LDLR and PCSK9 (−259.85 to −303.21 kJ/mol) based on the docking score of the top 5 binding predictions (Supplemental Figure 8A). Besides, the ligand root mean square deviation, often used to evaluate the correctness of the docking geometry and optimal superimposition of the receptor-ligand binding, was lower in the binding of SDC4 and PCSK9 (39.12–85.16) compared with that of LDLR and PCSK9 (43.76–156.25) for the top 5 binding predictions, suggesting better binding between SDC4 and PCSK9 than in LDLR and PCSK9. Furthermore, co-immunoprecipitation analysis of mouse liver tissue lysates validated the occurrence of the predicted PCSK9-SDC4 binding in vivo (Supplemental Figure 8B).\\nRNA silencing validated SDC4 (syndecan-4) as a potential target of PCSK9 (proprotein convertase subtilisin/kexin 9)-induced pro-inflammatory response in macrophages, identified via hyperedge entanglement prediction (HEP) prediction. A, SDC4 mRNA level was measured in murine peritoneal macrophages of Ldlr-/- mice 1 week after intravenous injection with AAV-LacZ or AAV-PCSK9 (n=7 per group). B, mRNA levels of SDC4, IL-1b, NLRP3, CD40, and ICAM1 were measured by RT-qPCR in mouse Ldlr-/- peritoneal macrophages after pretreatment with siSDC4 or siControl (Ctrl) for 48 hours and stimulation with 2.5 μg/mL recombinant mouse PCSK9 (MmPCSK9) for 3 hours (n=4–5). P value was calculated by 1-way ANOVA, followed by Bonferroni multiple comparison test. Error bars indicate ± SEM. Data are reported as mean ± SEM.\",\n \"We further validated our prediction that SDC4 is a potential target of PCSK9 using siRNA silencing of mouse Sdc4 in mouse peritoneal macrophages. Silencing Sdc4 with siSDC4 treatment achieved 93% reduction of SDC4 mRNA expression (Figure 6B). siSDC4 was then used to determine whether it could suppress PCSK9-induced pro-inflammatory molecules, including those predicted to be linked with PCSK9 by the HEP prediction (Figure 5C). siSDC4 suppressed mRNAs of the pro-inflammatory molecules IL-1β, NLRP3, CD40, and ICAM1, but not TNFα, MCP-1, TLR2, CXCL1, CXCL2, or CXCL10 after stimulation with recombinant PCSK9 (Figure 6B, Supplemental Figure 9). These data suggest that SDC4 partially regulates a PCSK9-induced pro-inflammatory response. Furthermore, these mRNA silencing results substantiate the validity of the prediction, yielding the integration of PCSK9 in our hyperedge of 24 disconnected differentially expressed transcripts.\",\n \"Our study demonstrated that circulating PCSK9 produced various changes in Ldlr-/- mice. We, however, attempted to examine the potential impact of local PCSK9 production in vein grafts. Immunohistochemical analysis using an anti-PCSK9 antibody in previous studies showed that vascular smooth muscle cells in human atherosclerotic plaque and collar-induced neointima of murine carotid artery express PCSK9.45,46 Based on these results, we examined whether locally produced PCSK9 in vein grafts may also contribute to vein graft lesion development. The local expression of PCSK9, as determined by in situ hybridization in the intima of vein grafts, was minimal compared with that in the liver (Figure 1B; Supplemental Figure 10A). RT-qPCR analysis further confirmed numerically significantly less PCSK9 mRNA expression in vein grafts than in the liver, the primary source of circulating PCSK9 (Supplemental Figure 10B). Moreover, we examined which cells within vein grafts can produce PCSK9 in humans by comparing PCSK9 mRNA levels in human saphenous vein endothelial cells, smooth muscle cells, and human primary macrophages derived from peripheral blood mononuclear cells. While we detected substantial PCSK9 expression in human saphenous vein smooth muscle cells, the level was significantly lower than that in HepG2 (liver) cells (Supplemental Figure 10C). Human primary macrophages and human saphenous vein endothelial cells contained minimal PCSK9 mRNA (Supplemental Figure 10C). Recent studies reported a modest increase of PCSK9 mRNA and protein levels in macrophages exposed to pro-inflammatory stimuli.39,47 We, therefore, examined PCSK9 mRNA expression in human primary macrophages treated with pro-inflammatory stimuli. Although LPS and IFNγ induced a pro-inflammatory response in the macrophages, as determined by TNFα mRNA expression (Supplemental Figures 10D and 10E, right graphs), PCSK9 mRNA expression did not increase after LPS or IFNγ (Supplemental Figures 10D and 10E, left graphs). These results indicate that smooth muscle cells (SMC), but not macrophages, may primarily produce PCSK9 in vein grafts, but to a much lower extent than the liver.\",\n \"Although our study focuses on macrophages, we examined the role of PCSK9 in the activation of endothelial cells and SMCs because other investigators reported potential interactions between PCSK9 and these cell types.46,48 Endothelial cell activation precedes vein graft lesion development.49 Due to the scarcity of mRNA in CD31-positive endothelial cells in vein grafts, we sorted them from the liver and lungs of Ldlr-/- mice injected with AAV-PCSK9 instead to examine the effects of PCSK9 on endothelial cell activation in an LDLR-independent fashion. AAV-PCSK9 did not increase mRNA levels of adhesion molecules, such as VCAM-1, ICAM-1, and E-selectin, in EC from either organ when compared with AAV-LacZ (Supplemental Figure 11A). We further found in vitro that recombinant PCSK9 did not increase mRNA levels of these adhesion molecules in human saphenous vein endothelial cells (Supplemental Figure 11B). These results indirectly indicate that PCSK9 may not induce endothelial activation in vein grafts.\\nSMC modulation, migration, and proliferation contribute to the development of vascular disorders, including atherosclerosis50 and vein graft.51 A recent study showed that PCSK9 may sustain SMC dedifferentiation, migration, and proliferation in neointimal hyperplasia in response to vascular injury,46 suggesting the involvement of SMCs in PCSK9-induced vein graft lesion development. To explore this possibility, we examined the effects of recombinant PCSK9 on SMC dedifferentiation, migration, and proliferation in human saphenous vein smooth muscle cells. Recombinant PCSK9 slightly decreased SM22α mRNA levels (0.74 and 0.68-fold for 1.0 and 5.0 µg/mL, respectively), but did not affect differentiation molecules, such as αSMA, calponin, SM1, or SM2 (Supplemental Figure 12A). In addition, recombinant PCSK9 did not affect PDGF-induced migration activity of human saphenous vein smooth muscle cells (Supplemental Figure 12B). A BrdU assay using human saphenous vein smooth muscle cells showed that recombinant PCSK9 did not enhance 10% fetal bovine serum-induced proliferation (Supplemental Figure 12C). AAV-PCSK9 did not affect the SMC content in vein grafts (Supplemental Figure 12D). Collectively, in our experimental settings, our data did not provide the mechanistic evidence in vitro and in vivo that activation of endothelial cells or SMCs contribute to PCSK9-induced vein graft lesion formation. These results indicate that circulating PCSK9 promotes vein graft lesion development predominantly through macrophage activation.\",\n \"This study investigated LDLR-independent mechanisms by which PCSK9 induces pro-inflammatory activation of macrophages and accelerates vein graft lesion development. The key findings include: (1) AAV-PCSK9 did not change lipid profile but promoted vein graft lesion development in Ldlr-/- mice, suggesting LDLR-independent mechanisms; (2) AAV-PCSK9 increased macrophage accumulation and decreased fibrillar collagen in vein graft lesions; (3) PCSK9 induced macrophage activation in Ldlr-/- mice; (4) unbiased global transcriptomics of Ldlr-/- mouse macrophages stimulated with recombinant PCSK9 revealed activation of several pro-inflammatory pathways; (5) a combination of the global transcriptomics and novel network-based hyperedge entanglement prediction analysis showed NF-κB signaling pathway and OLR1(LOX-1) could play a key role in vein graft lesion development; (6) SDC4, a heparan sulfate proteoglycan, could be a novel target of PCSK9 that mediates pro-inflammatory responses in macrophages.\\nReversed autologous saphenous veins are widely used for surgical bypass for treatment of lower extremity peripheral artery disease and coronary artery disease. Peripheral artery disease affects approximately 8.5 million people in the United States aged ≥40 years52 and a total of 202 million worldwide, as of 2010.53 The incidence of peripheral artery disease has been rising by 23.5% globally from 2000 to 2010 largely due to aging populations.53 Of note, 40% of vein grafts for peripheral artery disease are failing or fail within a year.1 The volume of CABG procedures for severe coronary artery disease has declined since 1998; however, 371 000 bypass procedures were still performed in the United States in 2014,52 and approximately 30% of these costly procedures are failing or fail 12 to 18 months after surgery.2\\nEvidence suggests that pro-inflammatory responses to PCSK9 in macrophages and arterial atherosclerotic lesions may primarily depend on LDLR. Recombinant PCSK9 induced TNFα mRNA in bone marrow-derived macrophages mainly in an LDLR-dependent fashion.19 Furthermore, overexpression of human PCSK9 in macrophages promoted atherosclerotic lesions.28 Deletion of the PCSK9 gene in the liver reduced atherosclerotic lesions, primarily via mechanisms dependent on LDLR.29 To the best of our knowledge, no in vivo studies have demonstrated LDLR-independent pro-inflammatory effects of PCSK9 on vascular lesion development. The present study demonstrated that circulating PCSK9 induces macrophage activation and vein graft lesion development in LDLR-independent mechanisms. The relative contributions of the LDLR-dependent and independent mechanisms, however, remain unknown.\\nClinical reports demonstrated that cholesterol-lowering with PCSK9 inhibitors reverses the pro-inflammatory profile of monocytes in patients with hypercholesterolemia,54 indicating the lipid-dependent pro-inflammatory effects of PCSK9 in humans. Clinical trials also demonstrated that PCSK9 inhibitors lower not only serum levels of LDL cholesterol but also lipoprotein (a) (Lp(a)).55 Mice lack apolipoprotein (a) or Lp(a),56 which enabled us to rule out the role of Lp(a) in the effects of PCSK9 in our in vivo study. Although genetic evidence suggests LDL-dependent effects of PCSK9 in gain-of-function and loss-of-function carriers57 and the clinical benefits of antiPCSK9 strategies derive largely from LDL lowering, nonLDL-dependent effects might also contribute.58 As serum total cholesterol and triglyceride levels did not differ in AAV-LacZ and AAV-PCSK9 groups, our findings in Ldlr-/- mice do not seem to depend on PCSK9’s effects on the profile of circulating lipids.\\nHigh sensitivity C-reactive protein is the most established marker of inflammation to predict cardiovascular events independently in primary and secondary prevention.59 Clinical trials with PCSK9 inhibitors did not show any significant changes in high sensitivity C-reactive protein levels in patients with coronary artery disease.60,61 Antiinflammatory changes in monocyte phenotype with PCSK9 antibodies were not accompanied by a CRP (C-reactive protein) reduction in patients with familial hypercholesterolemia, suggesting that CRP may not solely capture changes in local inflammation produced by PCSK9 inhibition.54 In addition, the ATHEROREMO-IVUS study (The European Collaborative Project on Inflammation and Vascular Wall Remodeling in Atherosclerosis - Intravascular Ultrasound) demonstrated a linear relationship between serum levels of PCSK9 and the fraction or amount of the necrotic core in the nonculprit lesions of patients with acute coronary syndrome, independently of serum levels of LDL cholesterol. Our study provides support for the notion that circulating PCSK9 may exert pro-inflammatory effects on vascular lesions. Further studies are needed to explore whether PCSK9 inhibition ameliorates local inflammation in vascular lesions in patients.\\nIn general, multi-omics technologies, such as proteomics and transcriptomics, identify many differentially expressed molecules. It is challenging to assess direct interactions between these molecules or their causal roles in a biological or disease context by examining each molecule or interaction among many. This challenge has driven efforts in computational network biomedicine to develop unbiased prediction methods. By exploiting the mere topology of a protein-protein interaction network, one can quantify the likelihood of direct interactions between a molecule of interest (eg, PCSK9 in the present study) and a hyperedge of differentially expressed molecules (which represents a network functional module). We used a novel node2hyperedge entanglement method20 and showed any of the 24 differentially expressed transcripts, including NF-κB signaling molecules, OLR1(LOX-1), and SDC4, as a potential target of PCSK9 associated with inflammation. In the present study, we identified SDC4 as a potential novel target of PCSK9. Although Sdc4 was not a high-ranking differential expressed transcript, its fold change after stimulation with recombinant PCSK9 in the RNA-sequencing was statistically significant. We further performed in vitro experiments involving siRNA silencing in primary macrophages to substantiate our computational prediction platform and demonstrated that SDC4 indeed binds to PCSK9 in vivo and mediates PCSK9-induced pro-inflammatory responses. The contribution of macrophage SDC4 to PCSK9-induced pro-inflammatory responses in vivo remains to be elucidated. This novel prediction method may offer a powerful and necessary tool for the completion of the human interactome and for generating new hypotheses.\\nSDC4 regulates several robust pro-inflammatory molecules, such as IL-1β, NLRP3 (NLR family pyrin domain containing 3), CD40 (cluster of differentiation 40) and ICAM1. Various endogenous signals abundantly present in atherosclerotic lesions activates NLRP3, a key mediator of IL-1β maturation.62,63 T cell-mediated macrophage activation involves CD40 to produce interstitial collagens and tissue-factor pro-coagulant.64 The role of macrophage ICAM1 in atherosclerosis remains obscure, but a recent study reported that macrophage ICAM1 suppresses “M2-like” macrophages during tumor progression. SDC4 silencing inhibited these pro-inflammatory molecules. The magnitude of inhibition, however, was modest, suggesting that while we propose SDC4 is a novel target of PCSK9, but PCSK9 may also target other molecules (eg, NF-κB signaling molecules). The mechanisms mediated via SDC4 we report here appears to be one of multiple pathways affected by PCSK9. Heparan sulfate proteoglycans modulate macrophage activation by maintaining cells in a quiescent state through the capture and sequestration of IFNβ.65,66 We speculate that SDC4 serves as a mediator of PCSK9-induced pro-inflammatory responses through the capture of PCSK9 on the surface of macrophages as seen in the hepatocyte.43 Our data demonstrated that NF-κB may mediate pro-inflammatory responses downstream of the PCSK9 and SDC4 binding. Understanding more detailed mechanisms by which SDC4 mediates the pro-inflammatory responses and their downstream signaling induced by PCSK9 requires future investigations.\\nOur study does not demonstrate definitively that local PCSK9 plays an important role in vein graft lesion development. In situ hybridization and qPCR data suggest that SMCs may be the primary source of PCSK9 in these lesions. The expression level of PCSK9 in SMCs was much lower than in the liver, although the relative gene expression level may not be a valid determinant of its biological role or importance. Whether and how circulating PCSK9 affects lesional macrophages through the contact with endothelial cells (which in turn affects macrophage phenotype or the entry into the subendothelial space, which results in direct contact with macrophages) remains to be elucidated.67 Future work needs to define detailed molecular mechanisms.\\nOur data do not demonstrate that human primary macrophages can express PCSK9, even when stimulated with pro-inflammatory stimuli, which was consistent with previous studies.45,68 Evidence suggests, however, the participation of endogenous PCSK9 in macrophages during pro-inflammatory responses. In vitro experiments showed that overexpression of PCSK9 in macrophage-like THP-1 cells enhanced the pro-inflammatory response induced by oxidized LDL via NF-κB activation.38 In contrast, PCSK9 gene silencing inhibited the response.38,47 In vivo evidence also showed that overexpression of human PCSK9 in macrophages increased atherosclerotic lesion formation in Apoe-/- mice.28 These results suggest that not only circulating PCSK9 produced from the liver but also local PCSK9 produced from SMCs and macrophages may induce a pro-inflammatory response in macrophages and contribute to vascular lesion development. What cell types other than SMCs contribute to monocyte/macrophage infiltration, and activation remains to be elucidated.\\nThe present study demonstrates several lines of novel evidence that circulating PCSK9 promotes macrophage activation and vein graft lesion development via mechanisms independent of blood cholesterol and LDLR degradation. Despite current therapeutics and research advancements, effective medical solutions are limited to prevent vein graft failure. Our study provides molecular bases that circulating PCSK9 deserves investigations as a potential therapeutic target for vein graft failure.\",\n \"[SUBTITLE] Acknowledgments [SUBSECTION] We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\\nWe thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\\n[SUBTITLE] Sources of Funding [SUBSECTION] This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\\nThis work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\\n[SUBTITLE] Disclosures [SUBSECTION] Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\\nPfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\\n[SUBTITLE] Supplemental Materials [SUBSECTION] Online Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\\nOnline Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\",\n \"We thank Ryo Kawakami, Yuki Tsukano, Dayanna C. Romero, Whitney S. Golden, Keishi Nihira, Andrew K. Mlynarchik, Edward Guzman, and Daniel G. Anderson for their technical assistance. The graphical figure was created with Biorender.com. Additional Information: Coauthor Amitabh Sharma died on November 24, 2019.\",\n \"This work was supported by the National Institute of Health grants R01HL126901 and R01HL149302, and Pfizer ASPIRE Award to MA; R01HL136431, R01HL147095, and R01HL141917 to EA; and the MSD Scholarships for Overseas Study, and Japan Society for the Promotion of Science Fellowships for Overseas Researchers to SK.\",\n \"Pfizer was not involved in the study other than funding. MA also received research grants from Kowa Company and Sanofi.\",\n \"Online Supplemental Materials and Methods\\nSupplemental Figures 1–14\\nSupplemental Tables 1–4\\nReferences 69–75\",\n \"\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,"methods","data-availability","methods",null,null,null,null,"results",null,null,null,null,null,null,null,null,"discussion",null,null,null,"COI-Statement",null,"supplementary-material"],"string":"[\n null,\n null,\n \"methods\",\n \"data-availability\",\n \"methods\",\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n \"COI-Statement\",\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["graft occlusion, vascular","inflammation","macrophage activation","receptors, LDL","systems biology"],"string":"[\n \"graft occlusion, vascular\",\n \"inflammation\",\n \"macrophage activation\",\n \"receptors, LDL\",\n \"systems biology\"\n]"}}},{"rowIdx":360,"cells":{"title":{"kind":"string","value":"Optimizing the implementation of lung cancer screening in Scotland: Focus group participant perspectives in the LUNGSCOT study."},"pmid":{"kind":"string","value":"36263948"},"background_abstract":{"kind":"string","value":"Targeted lung cancer screening is effective in reducing lung cancer and all-cause mortality according to major trials in the United Kingdom and Europe. However, the best ways of implementing screening in local communities requires an understanding of the population the programme will serve. We undertook a study to explore the views of those potentially eligible for, and to identify potential barriers and facilitators to taking part in, lung screening, to inform the development of a feasibility study."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"Men and women aged 45-70, living in urban and rural Scotland, and either self-reported people who smoke or who recently quit, were invited to take part in the study via research agency Taylor McKenzie. Eleven men and 14 women took part in three virtual focus groups exploring their views on lung screening. Focus group transcripts were transcribed and analysed using thematic analysis, assisted by QSR NVivo."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Three overarching themes were identified: (1) Knowledge, awareness and acceptability of lung screening, (2) Barriers and facilitators to screening and (3) Promoting screening and implementation ideas. Participants were largely supportive of lung screening in principle and described the importance of the early detection of cancer. Emotional and psychological concerns as well as system-level and practical issues were discussed as posing barriers and facilitators to lung screening."},"results_abstract_label":{"kind":"string","value":"FINDINGS"},"conclusions_abstract":{"kind":"string","value":"Understanding the views of people potentially eligible for a lung health check can usefully inform the development of a further study to test the feasibility and acceptability of lung screening in Scotland."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Male","Humans","Female","Early Detection of Cancer","Focus Groups","Lung Neoplasms","Mass Screening","Scotland","Qualitative Research"],"string":"[\n \"Male\",\n \"Humans\",\n \"Female\",\n \"Early Detection of Cancer\",\n \"Focus Groups\",\n \"Lung Neoplasms\",\n \"Mass Screening\",\n \"Scotland\",\n \"Qualitative Research\"\n]"},"pmcid":{"kind":"string","value":"9700133"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"We conducted three focus groups with self‐reported ‘heavy smokers’ living in rural, urban and deprived areas of Scotland to ascertain their views on barriers to lung screening. Focus groups were considered the most appropriate method for engaging with those eligible for screening, as they enabled an understanding of how participants perceive the prospect of lung screening through discussion and they enable the sharing and development of ideas.\n19\n Findings from the focus groups will feed into the development and design of the lung screening pilot. We employed a range of strategies and reflexive practices to ensure that focus groups were participant‐led and data‐driven.\n20\n\n\n[SUBTITLE] Identifying participants [SUBSECTION] Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\nParticipants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\n[SUBTITLE] Recruitment and sampling [SUBSECTION] We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\nWe aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\n[SUBTITLE] Focus groups and consent [SUBSECTION] Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\nThree focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\n[SUBTITLE] Analysis [SUBSECTION] Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\nFocus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\n[SUBTITLE] Patient advisory group [SUBSECTION] The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\nThe patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks."},"results_title":{"kind":"null"},"results_text":{"kind":"null"},"conclusions_title":{"kind":"string","value":"CONCLUSIONS"},"conclusions_text":{"kind":"string","value":"This focus group study has already identified several perceived individual, practical and system barriers and facilitators to participation in a pilot lung screening programme in Scotland using LDCT. While our results resonate with existing literature in this field, they will be helpful in addressing factors which are especially important in Scotland if it is to embrace lung cancer screening—reducing health inequalities, engaging deprived populations and ensuring access in remote and rural areas. The findings will inform the design and implementation of a Scottish pilot lung screening study."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Identifying participants","Recruitment and sampling","Focus groups and consent","Analysis","Patient advisory group","FINDINGS","Knowledge, awareness and acceptability of lung screening","Barriers and facilitators to screening","Individual level influences on screening intentions","Psychological and emotional concerns","Perceived risk and fatalism","Stigma and judgement as a barrier","Mistrust of healthcare professionals and services","Practical and system barriers","Promoting screening and implementation ideas","Managing fears and expectations","Improving accessibility","Summary","Comparison with existing literature and theory","Strengths and limitations","Implications and future work","AUTHOR CONTRIBUTIONS","ETHICS STATEMENT"],"string":"[\n \"INTRODUCTION\",\n \"Identifying participants\",\n \"Recruitment and sampling\",\n \"Focus groups and consent\",\n \"Analysis\",\n \"Patient advisory group\",\n \"FINDINGS\",\n \"Knowledge, awareness and acceptability of lung screening\",\n \"Barriers and facilitators to screening\",\n \"Individual level influences on screening intentions\",\n \"Psychological and emotional concerns\",\n \"Perceived risk and fatalism\",\n \"Stigma and judgement as a barrier\",\n \"Mistrust of healthcare professionals and services\",\n \"Practical and system barriers\",\n \"Promoting screening and implementation ideas\",\n \"Managing fears and expectations\",\n \"Improving accessibility\",\n \"Summary\",\n \"Comparison with existing literature and theory\",\n \"Strengths and limitations\",\n \"Implications and future work\",\n \"AUTHOR CONTRIBUTIONS\",\n \"ETHICS STATEMENT\"\n]"},"other_sections_texts":{"kind":"list like","value":["Lung cancer remains one of the major causes of cancer mortality globally, and Scotland has high incidence and low survival rates compared with the rest of Europe.\n1\n, \n2\n, \n3\n, \n4\n Despite many initiatives to raise awareness and encourage early symptomatic presentation, most lung cancers are diagnosed at a late stage, and overall survival is poor.\n5\n A recent Public Health Scotland report revealed that the Covid‐19 pandemic has had a further negative impact on the clinical presentation of lung cancer, with a reduction in diagnoses during the lockdown periods, and an anticipated wave of late diagnoses from this backlog.\n6\n\n\nOver the last 5 years, there has been a growing body of evidence from trials in the United States and continental Europe, and pilot studies in the United Kingdom, demonstrating survival gains from low‐dose computed tomography (LDCT) screening for lung cancer in high‐risk populations.\n7\n, \n8\n, \n9\n, \n10\n A recent meta‐analysis of this evidence provides a strong case for the implementation of targeted lung screening, supported by expert clinical opinion,\n11\n and the UK National Screening Committee has recently recommended in favour of targeted lung screening. However, there are still many lung screening implementation challenges to be addressed.\n12\n, \n13\n Importantly, lung screening is most effective if targeted at high‐risk groups—ideally using validated risk prediction tools to assess who is at the highest risk of developing lung cancer and would most benefit from screening, namely smokers or recent quitters, aged 55–74 years.\nTo date, UK pilots and trials, including the Early Detection of Cancer of the Lung Scotland (ECLS) trial in Scotland using blood biomarkers to detect early signs of lung cancer, have shown variable and socially patterned uptake of screening,\n8\n, \n14\n, \n15\n, \n16\n and work has been done to understand barriers to participation. Uptake remains lower in more marginalized groups—that is, heavy smokers living in more deprived areas.\n17\n Further research to understand the views of Scotland's population on the acceptability of lung screening, and barriers and facilitators (related to issues of geography, rural and urban deprivation and a high burden of multimorbidity), will help shape future pilots and programmes in Scotland.\nThis focus group study explores the views of people potentially eligible for lung screening, identifies perceived barriers and facilitators to participation, and examines strategies to optimize the implementation of lung screening in Scotland. It forms part of the LUNGSCOT study, which is examining the feasibility of introducing lung screening in Scotland.\n18\n\n","Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.","We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.","Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.","Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.","The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.","Eleven females and 14 males aged 45–70 years living in a mix of urban and rural areas in Scotland took part across the three focus groups. Eleven participants were current smokers, and 14 had quit within the previous 2 years. All participants were from the lower socio‐economic grades (SEG): C2 (skilled manual workers), D (semiskilled and unskilled manual workers) and E (nonworking). Twenty‐one participants were White British/Scottish, one person was Black British, one British Asian and one South Asian. All participants had school‐level or vocational qualifications but no one had a higher education degree. See Tables 1, 2, 3 for participant characteristics.\nFocus Group 1 participant characteristics\n\nNote: Four males, four females, age 46–71 years, from six different health board regions, living in urban areas. Four current smokers and four of whom had quit within the last 2 years.\nFocus Group 2 participant characteristics\n\nNote: Four females, three males, age range 45–62 years, from three different health board regions, living in rural areas, three smokers and four of whom have quit within the last 2 years.\nFocus Group 3 participant characteristics\n\nNote: Three females, seven males, age range 45–68 years, from three different health board regions, living in urban areas, four smokers and six of whom have quit within the last 2 years.\nOur analysis identified three overarching themes in the data: (1) Knowledge, awareness and acceptability of lung screening, (2) Barriers and facilitators to screening and (3) Promoting screening and implementation ideas.\n[SUBTITLE] Knowledge, awareness and acceptability of lung screening [SUBSECTION] There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\nThere is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\n[SUBTITLE] Barriers and facilitators to screening [SUBSECTION] It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nIt was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\n[SUBTITLE] Promoting screening and implementation ideas [SUBSECTION] There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nThere was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.","It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.","On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).","The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)","Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)","A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)","A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).","A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.","There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)","In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.","There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\n26\n, \n27\n, \n28\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\n27\n, \n29\n, \n30\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\n17\n, \n31\n\n\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\n28\n, \n32\n, \n33\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\n29\n, \n34\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\n35\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\n36\n, \n37\n, \n38\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\n28\n, \n33\n, \n39\n, \n40\n Risk and decision‐making are discussed further below.\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\n31\n, \n38\n, \n41\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\n42\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\n17\n, \n29\n, \n38\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\n17\n, \n43\n, \n44\n, \n45\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\n46\n, \n47\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\n36\n as well as being central to the long‐term cost‐effectiveness of screening.\n13\n\n\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\n17\n, \n32\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\n48\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\n48\n, \n49\n, \n50\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\n51\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\n7\n, \n8\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\n52\n, \n53\n\n\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\n27\n, \n32\n, \n46\n\n\nVon Wagner et al.\n50\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\n54\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\n55\n, \n56\n, \n57\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\n37\n, \n39\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\n48\n, \n49\n\n\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\n58\n, \n59\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\n12\n, \n60\n\n","This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\n61\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.","This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\n62\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\n40\n, \n45\n, \n46\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\n59\n, \n63\n, \n64\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\n\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\n\nMinimize steps in the screening process to lower opportunities for delays and associated distress\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\nIncorporating discussion of fears associated with screening into information materials\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\nEnsuring a timely and sensitive approach to smoking cessation advice","All authors apart from Jasmin Rostron and Lynsey R. Brown contributed toward the design of the focus groups and helped develop study documents. Debbie Cavers and Mia Nelson conducted the focus groups. Debbie Cavers, Mia Nelson and Jasmin Rostron coded and analysed the transcripts. Lynsey R. Brown read and commented on the analysis. Debbie Cavers drafted the manuscript and all authors edited and refined subsequent drafts.","The LUNGSCOT focus group study was approved by the University of Edinburgh Medical School Ethics Committee on 19 March 2021, reference 21‐EMREC‐002. All participants gave consent before taking part in the focus groups."],"string":"[\n \"Lung cancer remains one of the major causes of cancer mortality globally, and Scotland has high incidence and low survival rates compared with the rest of Europe.\\n1\\n, \\n2\\n, \\n3\\n, \\n4\\n Despite many initiatives to raise awareness and encourage early symptomatic presentation, most lung cancers are diagnosed at a late stage, and overall survival is poor.\\n5\\n A recent Public Health Scotland report revealed that the Covid‐19 pandemic has had a further negative impact on the clinical presentation of lung cancer, with a reduction in diagnoses during the lockdown periods, and an anticipated wave of late diagnoses from this backlog.\\n6\\n\\n\\nOver the last 5 years, there has been a growing body of evidence from trials in the United States and continental Europe, and pilot studies in the United Kingdom, demonstrating survival gains from low‐dose computed tomography (LDCT) screening for lung cancer in high‐risk populations.\\n7\\n, \\n8\\n, \\n9\\n, \\n10\\n A recent meta‐analysis of this evidence provides a strong case for the implementation of targeted lung screening, supported by expert clinical opinion,\\n11\\n and the UK National Screening Committee has recently recommended in favour of targeted lung screening. However, there are still many lung screening implementation challenges to be addressed.\\n12\\n, \\n13\\n Importantly, lung screening is most effective if targeted at high‐risk groups—ideally using validated risk prediction tools to assess who is at the highest risk of developing lung cancer and would most benefit from screening, namely smokers or recent quitters, aged 55–74 years.\\nTo date, UK pilots and trials, including the Early Detection of Cancer of the Lung Scotland (ECLS) trial in Scotland using blood biomarkers to detect early signs of lung cancer, have shown variable and socially patterned uptake of screening,\\n8\\n, \\n14\\n, \\n15\\n, \\n16\\n and work has been done to understand barriers to participation. Uptake remains lower in more marginalized groups—that is, heavy smokers living in more deprived areas.\\n17\\n Further research to understand the views of Scotland's population on the acceptability of lung screening, and barriers and facilitators (related to issues of geography, rural and urban deprivation and a high burden of multimorbidity), will help shape future pilots and programmes in Scotland.\\nThis focus group study explores the views of people potentially eligible for lung screening, identifies perceived barriers and facilitators to participation, and examines strategies to optimize the implementation of lung screening in Scotland. It forms part of the LUNGSCOT study, which is examining the feasibility of introducing lung screening in Scotland.\\n18\\n\\n\",\n \"Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\",\n \"We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\\n21\\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\\nInclusion criteria\\n\\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\\n\\nMen and women\\nAge 45–70 (inclusive)\\nCurrent residence in Scotland\\nSelf‐identify as smoker or recent quitter (within the last 2 years)\\nAble to undertake focus group interview in English\\nWilling to discuss their views on lung screening\\nExclusion criteria\\n\\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\\n\\nLacking capacity to give informed consent\\nNever smoker\\nSmoker who quit more than 2 years ago\\nNon‐English speakers, preventing them from comfortably taking part in a discussion\\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\",\n \"Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\",\n \"Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\\n22\\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\\n19\\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\\n23\\n, \\n24\\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\",\n \"The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\",\n \"Eleven females and 14 males aged 45–70 years living in a mix of urban and rural areas in Scotland took part across the three focus groups. Eleven participants were current smokers, and 14 had quit within the previous 2 years. All participants were from the lower socio‐economic grades (SEG): C2 (skilled manual workers), D (semiskilled and unskilled manual workers) and E (nonworking). Twenty‐one participants were White British/Scottish, one person was Black British, one British Asian and one South Asian. All participants had school‐level or vocational qualifications but no one had a higher education degree. See Tables 1, 2, 3 for participant characteristics.\\nFocus Group 1 participant characteristics\\n\\nNote: Four males, four females, age 46–71 years, from six different health board regions, living in urban areas. Four current smokers and four of whom had quit within the last 2 years.\\nFocus Group 2 participant characteristics\\n\\nNote: Four females, three males, age range 45–62 years, from three different health board regions, living in rural areas, three smokers and four of whom have quit within the last 2 years.\\nFocus Group 3 participant characteristics\\n\\nNote: Three females, seven males, age range 45–68 years, from three different health board regions, living in urban areas, four smokers and six of whom have quit within the last 2 years.\\nOur analysis identified three overarching themes in the data: (1) Knowledge, awareness and acceptability of lung screening, (2) Barriers and facilitators to screening and (3) Promoting screening and implementation ideas.\\n[SUBTITLE] Knowledge, awareness and acceptability of lung screening [SUBSECTION] There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\\nThere is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\\n[SUBTITLE] Barriers and facilitators to screening [SUBSECTION] It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nIt was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\n[SUBTITLE] Promoting screening and implementation ideas [SUBSECTION] There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nThere was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\",\n \"It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\",\n \"On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\",\n \"The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\",\n \"Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\",\n \"A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\",\n \"A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\",\n \"A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\",\n \"There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\",\n \"In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\",\n \"There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\\n26\\n, \\n27\\n, \\n28\\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\\n27\\n, \\n29\\n, \\n30\\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\\n17\\n, \\n31\\n\\n\\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\\n28\\n, \\n32\\n, \\n33\\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\\n29\\n, \\n34\\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\\n35\\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\\n36\\n, \\n37\\n, \\n38\\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\\n28\\n, \\n33\\n, \\n39\\n, \\n40\\n Risk and decision‐making are discussed further below.\\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\\n31\\n, \\n38\\n, \\n41\\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\\n42\\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\\n17\\n, \\n29\\n, \\n38\\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\\n17\\n, \\n43\\n, \\n44\\n, \\n45\\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\\n46\\n, \\n47\\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\\n36\\n as well as being central to the long‐term cost‐effectiveness of screening.\\n13\\n\\n\\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\\n17\\n, \\n32\\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\\n48\\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\\n48\\n, \\n49\\n, \\n50\\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\\n51\\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\\n7\\n, \\n8\\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\\n52\\n, \\n53\\n\\n\\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\\n27\\n, \\n32\\n, \\n46\\n\\n\\nVon Wagner et al.\\n50\\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\\n54\\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\\n55\\n, \\n56\\n, \\n57\\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\\n37\\n, \\n39\\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\\n48\\n, \\n49\\n\\n\\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\\n58\\n, \\n59\\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\\n12\\n, \\n60\\n\\n\",\n \"This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\\n61\\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\",\n \"This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\\n62\\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\\n40\\n, \\n45\\n, \\n46\\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\\n59\\n, \\n63\\n, \\n64\\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distress\\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\\nIncorporating discussion of fears associated with screening into information materials\\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\\nEnsuring a timely and sensitive approach to smoking cessation advice\",\n \"All authors apart from Jasmin Rostron and Lynsey R. Brown contributed toward the design of the focus groups and helped develop study documents. Debbie Cavers and Mia Nelson conducted the focus groups. Debbie Cavers, Mia Nelson and Jasmin Rostron coded and analysed the transcripts. Lynsey R. Brown read and commented on the analysis. Debbie Cavers drafted the manuscript and all authors edited and refined subsequent drafts.\",\n \"The LUNGSCOT focus group study was approved by the University of Edinburgh Medical School Ethics Committee on 19 March 2021, reference 21‐EMREC‐002. All participants gave consent before taking part in the focus groups.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Identifying participants","Recruitment and sampling","Focus groups and consent","Analysis","Patient advisory group","FINDINGS","Knowledge, awareness and acceptability of lung screening","Barriers and facilitators to screening","Individual level influences on screening intentions","Psychological and emotional concerns","Perceived risk and fatalism","Stigma and judgement as a barrier","Mistrust of healthcare professionals and services","Practical and system barriers","Promoting screening and implementation ideas","Managing fears and expectations","Improving accessibility","DISCUSSION","Summary","Comparison with existing literature and theory","Strengths and limitations","Implications and future work","CONCLUSIONS","AUTHOR CONTRIBUTIONS","CONFLICT OF INTEREST","ETHICS STATEMENT","Supporting information"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Identifying participants\",\n \"Recruitment and sampling\",\n \"Focus groups and consent\",\n \"Analysis\",\n \"Patient advisory group\",\n \"FINDINGS\",\n \"Knowledge, awareness and acceptability of lung screening\",\n \"Barriers and facilitators to screening\",\n \"Individual level influences on screening intentions\",\n \"Psychological and emotional concerns\",\n \"Perceived risk and fatalism\",\n \"Stigma and judgement as a barrier\",\n \"Mistrust of healthcare professionals and services\",\n \"Practical and system barriers\",\n \"Promoting screening and implementation ideas\",\n \"Managing fears and expectations\",\n \"Improving accessibility\",\n \"DISCUSSION\",\n \"Summary\",\n \"Comparison with existing literature and theory\",\n \"Strengths and limitations\",\n \"Implications and future work\",\n \"CONCLUSIONS\",\n \"AUTHOR CONTRIBUTIONS\",\n \"CONFLICT OF INTEREST\",\n \"ETHICS STATEMENT\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Lung cancer remains one of the major causes of cancer mortality globally, and Scotland has high incidence and low survival rates compared with the rest of Europe.\n1\n, \n2\n, \n3\n, \n4\n Despite many initiatives to raise awareness and encourage early symptomatic presentation, most lung cancers are diagnosed at a late stage, and overall survival is poor.\n5\n A recent Public Health Scotland report revealed that the Covid‐19 pandemic has had a further negative impact on the clinical presentation of lung cancer, with a reduction in diagnoses during the lockdown periods, and an anticipated wave of late diagnoses from this backlog.\n6\n\n\nOver the last 5 years, there has been a growing body of evidence from trials in the United States and continental Europe, and pilot studies in the United Kingdom, demonstrating survival gains from low‐dose computed tomography (LDCT) screening for lung cancer in high‐risk populations.\n7\n, \n8\n, \n9\n, \n10\n A recent meta‐analysis of this evidence provides a strong case for the implementation of targeted lung screening, supported by expert clinical opinion,\n11\n and the UK National Screening Committee has recently recommended in favour of targeted lung screening. However, there are still many lung screening implementation challenges to be addressed.\n12\n, \n13\n Importantly, lung screening is most effective if targeted at high‐risk groups—ideally using validated risk prediction tools to assess who is at the highest risk of developing lung cancer and would most benefit from screening, namely smokers or recent quitters, aged 55–74 years.\nTo date, UK pilots and trials, including the Early Detection of Cancer of the Lung Scotland (ECLS) trial in Scotland using blood biomarkers to detect early signs of lung cancer, have shown variable and socially patterned uptake of screening,\n8\n, \n14\n, \n15\n, \n16\n and work has been done to understand barriers to participation. Uptake remains lower in more marginalized groups—that is, heavy smokers living in more deprived areas.\n17\n Further research to understand the views of Scotland's population on the acceptability of lung screening, and barriers and facilitators (related to issues of geography, rural and urban deprivation and a high burden of multimorbidity), will help shape future pilots and programmes in Scotland.\nThis focus group study explores the views of people potentially eligible for lung screening, identifies perceived barriers and facilitators to participation, and examines strategies to optimize the implementation of lung screening in Scotland. It forms part of the LUNGSCOT study, which is examining the feasibility of introducing lung screening in Scotland.\n18\n\n","We conducted three focus groups with self‐reported ‘heavy smokers’ living in rural, urban and deprived areas of Scotland to ascertain their views on barriers to lung screening. Focus groups were considered the most appropriate method for engaging with those eligible for screening, as they enabled an understanding of how participants perceive the prospect of lung screening through discussion and they enable the sharing and development of ideas.\n19\n Findings from the focus groups will feed into the development and design of the lung screening pilot. We employed a range of strategies and reflexive practices to ensure that focus groups were participant‐led and data‐driven.\n20\n\n\n[SUBTITLE] Identifying participants [SUBSECTION] Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\nParticipants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\n[SUBTITLE] Recruitment and sampling [SUBSECTION] We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\nWe aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\n[SUBTITLE] Focus groups and consent [SUBSECTION] Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\nThree focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\n[SUBTITLE] Analysis [SUBSECTION] Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\nFocus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\n[SUBTITLE] Patient advisory group [SUBSECTION] The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\nThe patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.","Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.","We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\n21\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\nInclusion criteria\n\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\n\nMen and women\nAge 45–70 (inclusive)\nCurrent residence in Scotland\nSelf‐identify as smoker or recent quitter (within the last 2 years)\nAble to undertake focus group interview in English\nWilling to discuss their views on lung screening\nExclusion criteria\n\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\n\nLacking capacity to give informed consent\nNever smoker\nSmoker who quit more than 2 years ago\nNon‐English speakers, preventing them from comfortably taking part in a discussion\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.","Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.","Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\n22\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\n19\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\n23\n, \n24\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.","The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.","Eleven females and 14 males aged 45–70 years living in a mix of urban and rural areas in Scotland took part across the three focus groups. Eleven participants were current smokers, and 14 had quit within the previous 2 years. All participants were from the lower socio‐economic grades (SEG): C2 (skilled manual workers), D (semiskilled and unskilled manual workers) and E (nonworking). Twenty‐one participants were White British/Scottish, one person was Black British, one British Asian and one South Asian. All participants had school‐level or vocational qualifications but no one had a higher education degree. See Tables 1, 2, 3 for participant characteristics.\nFocus Group 1 participant characteristics\n\nNote: Four males, four females, age 46–71 years, from six different health board regions, living in urban areas. Four current smokers and four of whom had quit within the last 2 years.\nFocus Group 2 participant characteristics\n\nNote: Four females, three males, age range 45–62 years, from three different health board regions, living in rural areas, three smokers and four of whom have quit within the last 2 years.\nFocus Group 3 participant characteristics\n\nNote: Three females, seven males, age range 45–68 years, from three different health board regions, living in urban areas, four smokers and six of whom have quit within the last 2 years.\nOur analysis identified three overarching themes in the data: (1) Knowledge, awareness and acceptability of lung screening, (2) Barriers and facilitators to screening and (3) Promoting screening and implementation ideas.\n[SUBTITLE] Knowledge, awareness and acceptability of lung screening [SUBSECTION] There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\nThere is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\n[SUBTITLE] Barriers and facilitators to screening [SUBSECTION] It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nIt was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\n[SUBTITLE] Promoting screening and implementation ideas [SUBSECTION] There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nThere was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\n25\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\n\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\n\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\n\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\n\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\nR1: Yeah, I 100 per cent agree with that.\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\n\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\n\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.","It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.","On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).","The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\n\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\n\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\n\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)","Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\n\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\n\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\n\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)","A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\n\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\n\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)","A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\n\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).","A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\n\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\n\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.","There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\n\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\n\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)","In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\n\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\n\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\n\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)","[SUBTITLE] Summary [SUBSECTION] This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\nThis focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\n[SUBTITLE] Comparison with existing literature and theory [SUBSECTION] There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\n26\n, \n27\n, \n28\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\n27\n, \n29\n, \n30\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\n17\n, \n31\n\n\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\n28\n, \n32\n, \n33\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\n29\n, \n34\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\n35\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\n36\n, \n37\n, \n38\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\n28\n, \n33\n, \n39\n, \n40\n Risk and decision‐making are discussed further below.\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\n31\n, \n38\n, \n41\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\n42\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\n17\n, \n29\n, \n38\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\n17\n, \n43\n, \n44\n, \n45\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\n46\n, \n47\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\n36\n as well as being central to the long‐term cost‐effectiveness of screening.\n13\n\n\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\n17\n, \n32\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\n48\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\n48\n, \n49\n, \n50\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\n51\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\n7\n, \n8\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\n52\n, \n53\n\n\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\n27\n, \n32\n, \n46\n\n\nVon Wagner et al.\n50\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\n54\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\n55\n, \n56\n, \n57\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\n37\n, \n39\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\n48\n, \n49\n\n\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\n58\n, \n59\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\n12\n, \n60\n\n\nThere is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\n26\n, \n27\n, \n28\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\n27\n, \n29\n, \n30\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\n17\n, \n31\n\n\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\n28\n, \n32\n, \n33\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\n29\n, \n34\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\n35\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\n36\n, \n37\n, \n38\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\n28\n, \n33\n, \n39\n, \n40\n Risk and decision‐making are discussed further below.\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\n31\n, \n38\n, \n41\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\n42\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\n17\n, \n29\n, \n38\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\n17\n, \n43\n, \n44\n, \n45\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\n46\n, \n47\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\n36\n as well as being central to the long‐term cost‐effectiveness of screening.\n13\n\n\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\n17\n, \n32\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\n48\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\n48\n, \n49\n, \n50\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\n51\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\n7\n, \n8\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\n52\n, \n53\n\n\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\n27\n, \n32\n, \n46\n\n\nVon Wagner et al.\n50\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\n54\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\n55\n, \n56\n, \n57\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\n37\n, \n39\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\n48\n, \n49\n\n\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\n58\n, \n59\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\n12\n, \n60\n\n\n[SUBTITLE] Strengths and limitations [SUBSECTION] This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\n61\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\nThis set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\n61\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\n[SUBTITLE] Implications and future work [SUBSECTION] This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\n62\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\n40\n, \n45\n, \n46\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\n59\n, \n63\n, \n64\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\n\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\n\nMinimize steps in the screening process to lower opportunities for delays and associated distress\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\nIncorporating discussion of fears associated with screening into information materials\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\nEnsuring a timely and sensitive approach to smoking cessation advice\nThis study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\n62\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\n40\n, \n45\n, \n46\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\n59\n, \n63\n, \n64\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\n\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\n\nMinimize steps in the screening process to lower opportunities for delays and associated distress\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\nIncorporating discussion of fears associated with screening into information materials\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\nEnsuring a timely and sensitive approach to smoking cessation advice","This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.","There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\n26\n, \n27\n, \n28\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\n27\n, \n29\n, \n30\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\n17\n, \n31\n\n\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\n28\n, \n32\n, \n33\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\n29\n, \n34\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\n35\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\n36\n, \n37\n, \n38\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\n28\n, \n33\n, \n39\n, \n40\n Risk and decision‐making are discussed further below.\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\n31\n, \n38\n, \n41\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\n42\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\n17\n, \n29\n, \n38\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\n17\n, \n43\n, \n44\n, \n45\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\n46\n, \n47\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\n36\n as well as being central to the long‐term cost‐effectiveness of screening.\n13\n\n\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\n17\n, \n32\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\n48\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\n48\n, \n49\n, \n50\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\n51\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\n7\n, \n8\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\n52\n, \n53\n\n\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\n27\n, \n32\n, \n46\n\n\nVon Wagner et al.\n50\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\n54\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\n55\n, \n56\n, \n57\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\n37\n, \n39\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\n48\n, \n49\n\n\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\n58\n, \n59\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\n12\n, \n60\n\n","This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\n61\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.","This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\n62\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\n40\n, \n45\n, \n46\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\n59\n, \n63\n, \n64\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\n\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\n\nMinimize steps in the screening process to lower opportunities for delays and associated distress\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\nIncorporating discussion of fears associated with screening into information materials\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\nEnsuring a timely and sensitive approach to smoking cessation advice","This focus group study has already identified several perceived individual, practical and system barriers and facilitators to participation in a pilot lung screening programme in Scotland using LDCT. While our results resonate with existing literature in this field, they will be helpful in addressing factors which are especially important in Scotland if it is to embrace lung cancer screening—reducing health inequalities, engaging deprived populations and ensuring access in remote and rural areas. The findings will inform the design and implementation of a Scottish pilot lung screening study.","All authors apart from Jasmin Rostron and Lynsey R. Brown contributed toward the design of the focus groups and helped develop study documents. Debbie Cavers and Mia Nelson conducted the focus groups. Debbie Cavers, Mia Nelson and Jasmin Rostron coded and analysed the transcripts. Lynsey R. Brown read and commented on the analysis. Debbie Cavers drafted the manuscript and all authors edited and refined subsequent drafts.","F. S. declares that the Universities of Dundee and St. Andrew's received funding from Oncimmune for the Early Detection of Cancer of the Lung Scotland (ECLS) trial from 2013 to 2021. The remaining authors declare no conflict of interest.","The LUNGSCOT focus group study was approved by the University of Edinburgh Medical School Ethics Committee on 19 March 2021, reference 21‐EMREC‐002. All participants gave consent before taking part in the focus groups.","Supporting information.\nClick here for additional data file."],"string":"[\n \"Lung cancer remains one of the major causes of cancer mortality globally, and Scotland has high incidence and low survival rates compared with the rest of Europe.\\n1\\n, \\n2\\n, \\n3\\n, \\n4\\n Despite many initiatives to raise awareness and encourage early symptomatic presentation, most lung cancers are diagnosed at a late stage, and overall survival is poor.\\n5\\n A recent Public Health Scotland report revealed that the Covid‐19 pandemic has had a further negative impact on the clinical presentation of lung cancer, with a reduction in diagnoses during the lockdown periods, and an anticipated wave of late diagnoses from this backlog.\\n6\\n\\n\\nOver the last 5 years, there has been a growing body of evidence from trials in the United States and continental Europe, and pilot studies in the United Kingdom, demonstrating survival gains from low‐dose computed tomography (LDCT) screening for lung cancer in high‐risk populations.\\n7\\n, \\n8\\n, \\n9\\n, \\n10\\n A recent meta‐analysis of this evidence provides a strong case for the implementation of targeted lung screening, supported by expert clinical opinion,\\n11\\n and the UK National Screening Committee has recently recommended in favour of targeted lung screening. However, there are still many lung screening implementation challenges to be addressed.\\n12\\n, \\n13\\n Importantly, lung screening is most effective if targeted at high‐risk groups—ideally using validated risk prediction tools to assess who is at the highest risk of developing lung cancer and would most benefit from screening, namely smokers or recent quitters, aged 55–74 years.\\nTo date, UK pilots and trials, including the Early Detection of Cancer of the Lung Scotland (ECLS) trial in Scotland using blood biomarkers to detect early signs of lung cancer, have shown variable and socially patterned uptake of screening,\\n8\\n, \\n14\\n, \\n15\\n, \\n16\\n and work has been done to understand barriers to participation. Uptake remains lower in more marginalized groups—that is, heavy smokers living in more deprived areas.\\n17\\n Further research to understand the views of Scotland's population on the acceptability of lung screening, and barriers and facilitators (related to issues of geography, rural and urban deprivation and a high burden of multimorbidity), will help shape future pilots and programmes in Scotland.\\nThis focus group study explores the views of people potentially eligible for lung screening, identifies perceived barriers and facilitators to participation, and examines strategies to optimize the implementation of lung screening in Scotland. It forms part of the LUNGSCOT study, which is examining the feasibility of introducing lung screening in Scotland.\\n18\\n\\n\",\n \"We conducted three focus groups with self‐reported ‘heavy smokers’ living in rural, urban and deprived areas of Scotland to ascertain their views on barriers to lung screening. Focus groups were considered the most appropriate method for engaging with those eligible for screening, as they enabled an understanding of how participants perceive the prospect of lung screening through discussion and they enable the sharing and development of ideas.\\n19\\n Findings from the focus groups will feed into the development and design of the lung screening pilot. We employed a range of strategies and reflexive practices to ensure that focus groups were participant‐led and data‐driven.\\n20\\n\\n\\n[SUBTITLE] Identifying participants [SUBSECTION] Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\\nParticipants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\\n[SUBTITLE] Recruitment and sampling [SUBSECTION] We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\\n21\\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\\nInclusion criteria\\n\\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\\n\\nMen and women\\nAge 45–70 (inclusive)\\nCurrent residence in Scotland\\nSelf‐identify as smoker or recent quitter (within the last 2 years)\\nAble to undertake focus group interview in English\\nWilling to discuss their views on lung screening\\nExclusion criteria\\n\\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\\n\\nLacking capacity to give informed consent\\nNever smoker\\nSmoker who quit more than 2 years ago\\nNon‐English speakers, preventing them from comfortably taking part in a discussion\\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\\nWe aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\\n21\\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\\nInclusion criteria\\n\\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\\n\\nMen and women\\nAge 45–70 (inclusive)\\nCurrent residence in Scotland\\nSelf‐identify as smoker or recent quitter (within the last 2 years)\\nAble to undertake focus group interview in English\\nWilling to discuss their views on lung screening\\nExclusion criteria\\n\\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\\n\\nLacking capacity to give informed consent\\nNever smoker\\nSmoker who quit more than 2 years ago\\nNon‐English speakers, preventing them from comfortably taking part in a discussion\\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\\n[SUBTITLE] Focus groups and consent [SUBSECTION] Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\\nThree focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\\n[SUBTITLE] Analysis [SUBSECTION] Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\\n22\\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\\n19\\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\\n23\\n, \\n24\\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\\nFocus group recordings were transcribed verbatim and subject to Braun and Clarke's\\n22\\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\\n19\\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\\n23\\n, \\n24\\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\\n[SUBTITLE] Patient advisory group [SUBSECTION] The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\\nThe patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\",\n \"Participants were recruited through Taylor McKenzie (TM), a Scottish‐based company that specializes in qualitative research, to identify members of the public eligible to take part in the focus groups (https://www.taylormckenzie.co.uk). TM developed a study‐specific screening questionnaire (further details below) to allow the purposive sampling of eligible participants from their extensive database of people willing to take part in health, social or marketing research.\",\n \"We aimed to recruit up to 24 people from across Scotland to take part in three separate focus groups of 8 people, considered to be an appropriate number to identify a range of views.\\n21\\n Interested people responded to recruitment notices posted on TM's mailing list and social media pages. We used the inclusion and exclusion criteria listed In Box 1.\\nInclusion criteria\\n\\nMen and womenAge 45–70 (inclusive)Current residence in ScotlandSelf‐identify as smoker or recent quitter (within the last 2 years)Able to undertake focus group interview in EnglishWilling to discuss their views on lung screening\\n\\nMen and women\\nAge 45–70 (inclusive)\\nCurrent residence in Scotland\\nSelf‐identify as smoker or recent quitter (within the last 2 years)\\nAble to undertake focus group interview in English\\nWilling to discuss their views on lung screening\\nExclusion criteria\\n\\nLacking capacity to give informed consentNever smokerSmoker who quit more than 2 years agoNon‐English speakers, preventing them from comfortably taking part in a discussion\\n\\nLacking capacity to give informed consent\\nNever smoker\\nSmoker who quit more than 2 years ago\\nNon‐English speakers, preventing them from comfortably taking part in a discussion\\nThose who responded were provided with a study information sheet by TM and given 7–14 days to consider taking part. TM drew up a list of eligible participants (based on information on their database, e.g., socioeconomic grade [SEG], occupation, and from speaking to potential participants directly, e.g., smoking status) and added it to a secure portal for the researchers to access and review. Eligible participants were allocated to a focus group at a preset date and time. Participants were offered a financial reimbursement for their time, paid via TM.\",\n \"Three focus groups, lasting approximately 75 min, were run virtually using the online video conferencing platform Zoom, selected due to likely participant familiarity with it, and to comply with the prevailing government restrictions on face‐to‐face meetings at the time due to the Covid‐19 pandemic. Participants were asked to sign a digital consent form via TM ahead of the focus group and verbal consent was agreed upon at the beginning of each focus group. Focus groups were led and facilitated by two researchers (D. C. and M. N.—health services researchers), with one group comprising those living in rural areas and the other two urban groups. Participants did not know one another before the focus group. The format and content of the focus groups were developed by a subgroup of the research team with expertise in behavioural aspects of cancer screening, drawing on relevant literature. The lung screening process was explained to participants: eligible people would be offered a LDCT scan to detect any lung conditions, one of which is lung cancer. The topic guide (see Supporting Information: 1) covered views on lung screening, with a particular focus on perceived barriers to taking part, personal resources to facilitate screening, understanding of the process, and input on what a good screening programme would look like. With consent, focus groups were recorded.\",\n \"Focus group recordings were transcribed verbatim and subject to Braun and Clarke's\\n22\\n thematic analysis, chosen as it is the most commonly used approach considered appropriate to derive key themes and ideas from the group discussions, taking context into account\\n19\\n and consistent with social constructionist underpinnings. Using thematic analysis allowed us to incorporate guidance specific to focus group interviewing and its impact on analysis, for example, considering dynamics, social comparison and power imbalances within groups.\\n23\\n, \\n24\\n Transcripts were read repeatedly and compared and contrasted to develop a set of common codes by D. C. These codes were applied back across the data and assigned to excerpts from the focus groups using QSR NVivo version 12 Pro (www.qsrinternational.com) by D. C. and J. R. (a research intern). Codes were further refined and a set of overarching themes and subthemes were inductively derived to interpret and explain the data, in discussion with the wider research team and patient advisory group. The LUNGSCOT team comprises health services researchers, health psychologists, clinicians, a health economist and patient and carer representatives. Themes were placed in the context of existing literature and theory to incorporate our findings into the wider evidence base.\",\n \"The patient advisory group was convened for the purpose of the wider LUNGSCOT study. The group comprises three patients and one carer with experience in lung cancer and two patients with other cancers. The group has been involved in the study design and commented on study documentation as well as two advisors reading transcripts and sharing their views on the analysis. The group meets quarterly to discuss study progress and opportunities to get involved in study tasks.\",\n \"Eleven females and 14 males aged 45–70 years living in a mix of urban and rural areas in Scotland took part across the three focus groups. Eleven participants were current smokers, and 14 had quit within the previous 2 years. All participants were from the lower socio‐economic grades (SEG): C2 (skilled manual workers), D (semiskilled and unskilled manual workers) and E (nonworking). Twenty‐one participants were White British/Scottish, one person was Black British, one British Asian and one South Asian. All participants had school‐level or vocational qualifications but no one had a higher education degree. See Tables 1, 2, 3 for participant characteristics.\\nFocus Group 1 participant characteristics\\n\\nNote: Four males, four females, age 46–71 years, from six different health board regions, living in urban areas. Four current smokers and four of whom had quit within the last 2 years.\\nFocus Group 2 participant characteristics\\n\\nNote: Four females, three males, age range 45–62 years, from three different health board regions, living in rural areas, three smokers and four of whom have quit within the last 2 years.\\nFocus Group 3 participant characteristics\\n\\nNote: Three females, seven males, age range 45–68 years, from three different health board regions, living in urban areas, four smokers and six of whom have quit within the last 2 years.\\nOur analysis identified three overarching themes in the data: (1) Knowledge, awareness and acceptability of lung screening, (2) Barriers and facilitators to screening and (3) Promoting screening and implementation ideas.\\n[SUBTITLE] Knowledge, awareness and acceptability of lung screening [SUBSECTION] There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\\nThere is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\\n[SUBTITLE] Barriers and facilitators to screening [SUBSECTION] It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nIt was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\n[SUBTITLE] Promoting screening and implementation ideas [SUBSECTION] There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nThere was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"There is currently no national lung screening programme in the United Kingdom, although parts of NHS England are offering lung screening with LDCT.\\n25\\n Participants were largely unaware of the concept of targeted lung screening:No, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\n\\nNo, I always thought that was something that happened if you develop, you know, or if they suspect you develop then you would have a check, otherwise nothing pre‐emptive…. (FG1, R5)\\nMost participants had heard of and participated in other forms of screening, including for breast, cervical and bowel cancer, and a few had been referred for lung checks for other reasons. They also described family members having had cancers picked up in this way, as well as their own experience of cancer:For me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\n\\nFor me, it really is, it's very important to be screened, especially breast, bowel, anything. I suffered myself from throat cancer ten years ago and I've been in remission for the past four years, so it's urgently important that people get this done, yeah. (FG3, R4)\\nWhen we described what lung screening would entail, participants were very supportive of this form of screening being available and welcomed the chance to have their lungs screened, with multiple participants saying it was ‘a great idea’. Participants were aware of the benefits of early (asymptomatic) detection and treatment:I think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\n\\nI think the screening is a good idea to catch things earlier or to see if somebody's got the disease or whatever that they didn't know they had. (FG1, R7)\\nParticipants also talked about the importance of screening early and not waiting until the ‘damage was done’, with discussions around age and smoking. It was suggested that raising awareness about lung cancer and lung screening should be introduced to school children, embedding knowledge of screening from an early age. Participants were also largely accepting of the fact that lung screening could pick up other issues,R3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.R1: Yeah, I 100 per cent agree with that.R4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\n\\nR3: Anything that shows up as a side‐line to it is a benefit and I think most people would welcome it.\\nR1: Yeah, I 100 per cent agree with that.\\nR4: Yeah, totally agree with that one as well, an added bonus. (FG3, R1, 3 and 4)\\nHowever, there were certain caveats and conditions to participating in screening. The concept of targeted screening for smokers was problematic for participants, as they discussed other risk factors for lung cancer apart from smoking and that some people who never smoke go on to develop lung cancer:There's other causes of lung cancer, it's not just smoking. (FG1, R7)\\n\\nThere's other causes of lung cancer, it's not just smoking. (FG1, R7)\\nParticipants introduced the potentially judgemental and stigmatizing nature of risk‐related eligibility to discussions. The importance of personal informed decision‐making and lack of coercion were also voiced by participants.\\nFacilitators asked participants about the role of smoking cessation advice in the lung screening process. Participants suggested that this would not put them off participating, although they reported a dislike of being pushed into stopping smoking or judged for their smoking, such that language and tone were important,I think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\n\\nI think you have to have that balance … for people to [not] think, ‘oh, we're going there and we're going to have that shoved down our throat [i.e., “forced on us”]’. It has to be choice, but I think always giving people the appropriate choices if somebody is ready to stop, […] and giving them the information being there available, but I don't think making it part of something, because what would then happen is people would then think they're getting this shoved down our throat, we're getting judged for smoking, […] So I think, yes, it's good to have all the information, but it's how it's given. (FG1, R6)\\nParticipants said that most people had received smoking cessation advice before, they already knew that smoking was harmful, and would not be offended by a health professional asking them about smoking. There was a strong sense among participants that the desire to quit smoking and action to quit was self‐motivated.\",\n \"It was clear that while there was strong support for the concept of lung screening with participants overwhelmingly in favour of it, a number of factors were raised which qualified their response—with the prospect of mediating the gap between reported intentions and performed screening behaviours in a real‐world scenario. Some of these same issues were potential motivators to participate in lung screening. Broadly, these can be represented under individual level and practical and system level factors.\\n[SUBTITLE] Individual level influences on screening intentions [SUBSECTION] On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nOn an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\n[SUBTITLE] Practical and system barriers [SUBSECTION] A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\\nA number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\",\n \"On an individual level, there were a number of cognitive, psychological and emotional factors influencing screening intentions.\\n[SUBTITLE] Psychological and emotional concerns [SUBSECTION] The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\nThe most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n[SUBTITLE] Perceived risk and fatalism [SUBSECTION] Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\nPerceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n[SUBTITLE] Stigma and judgement as a barrier [SUBSECTION] A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\nA common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n[SUBTITLE] Mistrust of healthcare professionals and services [SUBSECTION] A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\\nA number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\",\n \"The most pronounced psychological and emotional concerns reported were fear and worry about cancer and interactions with health services. Fear of invasive procedures, disruption of their lives, waiting for results and the challenges of facing a cancer diagnosis were off‐putting for a number of people:Maybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)It's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)A lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\n\\nMaybe you don't want to know, maybe you don't want to have cancer so it's better, you know, just to kind of blunder on and not find out, so not even to go and to be scared of going. Also if you find out you have something wrong then you're going to have to change your lifestyle to make things better. (FG1, R5)\\nIt's not knowing if you have the underlying issue or not, and then having to wait and then find out. I think that's maybe what puts a lot of people off not actually doing it. I think they're just maybe prepared, until they get a scare themselves and then go through a test, they're willing to just bypass it. (FG2, R5)\\nA lot of people are scared to come forward in case that the results of a test are positive. A lot of people don't want to know. And while they feel okay and there are no symptoms, that's fine. And then for someone to say to them, oh, by the way, you have this or that, it's quite scary for some people. (FG3, R3)\\nWhile participants (such as respondent 3 above) spoke of fear of the unknown, they were motivated to see what damage had been done to their lungs, showing the complexity of these thoughts in influencing behaviour,I wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\n\\nI wonder what stage my lungs are really at? And I'm sure other people think like that as well. […] It's like they know but we don't know and I think we should know. (FG3, R3)\\nOther issues raised in the focus groups related to problems engaging the older generation and males in particular, who they suggested were often stoical in their approach to health and illness and reluctant to burden health services—or they may see no point in screening when they are able to continue functioning. For some participants, the intention to be screened was related to being conscientious citizens and prioritizing one's health above other competing demands:You know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\\n\\nYou know, most people are concerned for their own health, they want to be healthy, they don't want to be a burden to the doctor or the NHS. But if you've got something wrong, early intervention is the answer. (FG3, R2)\",\n \"Perceived risk of lung cancer also appeared to influence people's screening intentions. Many people felt that their relatively young age and lack of symptoms meant that they were unlikely to have lung cancer and so screening was not relevant to them, though this was tied up with fear and a sense of fatalism in a complex way:I'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)I feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\n\\nI'm fit and I'm healthy. If I go and get it and it comes back it's chronic or it's terminal, fine, it's only terminal for as long as I'm going to last. (FG3, R4)\\nI feel myself slightly younger than everyone here that I see screening as a slightly older person's thing or a female thing, and you don't have to worry about it until you're like a certain age or something like that. (FG1, R5)\\nPerceived risk was also associated with family history and advancing age. Bad experiences of cancer in the family were discussed as both a source of avoidance of screening and a motivator to take part to avoid late detection and a poor prognosis:Yeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\n\\nYeah, I've been for a cervical one because unfortunately on my mum's side of the family it seems to run in the family. So, my mum and my older sister have all had hysterectomies at a young age … You have to go for these screenings, especially if you find it runs in your family. Even if it doesn't run in your family, the older we get, we are getting older and it's all inside as well so it's good to know what's going on. (FG2, R5)\\nAs above, fatalism was also evident in discussions among those who had smoked heavily and felt that lung cancer was unavoidable:At the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\\n\\nAt the end of the day, I've smoked since I was 15, so what damage is done [, is done]. (FG3, R4)\",\n \"A common theme across all focus groups was the role of judgement of smoking behaviour and perceived stigma as a barrier to presenting for lung screening:The stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\n\\nThe stigma of people who smoke is very real and they're made to feel uncomfortable. Although for years and years we were encouraged to do it, it was modern, it was yuppy, it was everything you wanted. It was cool, sophisticated, and yet now smokers feel … Well, I don't know, I'm only speaking for myself, I feel as if, oh well, it's your own fault, you caused it. I did cause it, but I was encouraged to cause it. (FG3, R4)\\nParticipants were also conscious of the cost of screening and did not want to be seen to take advantage of the system:No, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\\n\\nNo, but you could get members of the public being quite judgemental because I smoke so I know the health concerns, so if I'm choosing to do a risk‐taking behaviour, shall we say, I'm presuming this is something that the taxpayer is going to pay, so I'm thinking somebody might think why should I get tested for something that I'm putting myself at risk for. That might put people off. (FG1, R6)\",\n \"A number of participants reported good experiences with health services and a proactive approach to their health, driven by an awareness of the benefits of early detection. However, poor experiences with health services and healthcare professionals compromized trust and limited faith in services for some participants, and therefore an avoidance of any kind of interaction with them:I wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\n\\nI wouldn't trust my GP in that case to recommend, you know? Maybe he's biased, he's thinking, okay, that guy's smoking so he's just wasting time and money anyway so I won't recommend him. (FG1, R5)\\nParticipants in two of the groups discussed the fact that GPs are often overworked and a perception that patients' smoking results in GPs not investigating issues adequately or treating them fairly. By‐passing GPs and attending screening through an independent screening programme was seen as a positive thing. On the other hand, some participants reported good experiences and welcomed an endorsement from their GP, with one person commenting, ‘It's formal coming from your doctor’ (FG1, R5).\",\n \"A number of practical barriers to attending for screening were raised by participants. These included time off from work leading to loss of money, distance of travel to an appointment and access issues:If you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\n\\nIf you got a letter in saying, oh, you've got to go at ten o'clock in the morning and it's hardly worth going to work before that because you have to travel or whatever, so you might end up losing four hours' pay, you know what I mean? So if there's an incentive to encourage people for them not to lose money, I think you would get over 90 per cent of people would do it. (FG1, R3)\\nLiving in remote areas of Scotland, while having local access to primary care, was also voiced as an issue in terms of accessing secondary care, which could be a problem for a hospital‐based lung screening:We're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\n\\nWe're lucky in [place] that we've got a hospital and it covers the whole of the north but there are lots of places that … and there are a lot of old people that can't get to it. (FG1, R2)\\nParticipants also had some concerns about the ability of health services to meet the demands of screening in terms of providing an accessible and timely service. Delays and waiting for test results were a source of worry for people and there was a common reported perception of the NHS as an under‐resourced and over‐stretched service.\\nHaving outlined what the screening process would entail, participants did not voluntarily raise any concerns about radiation exposure, over‐diagnosis or invasiveness of the LDCT procedure. Participants were supportive of a ‘proper’ check of the lungs, including identification of nodules to be monitored or other incidental findings, particularly when they had existing lung problems.\",\n \"There was considerable discussion in each focus group on the acceptability and accessibility of screening, leading to suggestions of ways of promoting and implementing screening to increase participation. Participants' views on accessibility were of particular interest due to their socioeconomic status and geographical diversity. Suggestions related to the cognitive and psychological barriers to screening as well as practical issues.\\n[SUBTITLE] Managing fears and expectations [SUBSECTION] To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\nTo address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n[SUBTITLE] Improving accessibility [SUBSECTION] In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\nIn addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"To address fears around lung cancer, participants suggest that there is an emphasis in any information materials or advertising campaigns on positive messaging. Examples were given of being able to live to play with grandchildren and harnessing the successes of other screening programmes:I think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\n\\nI think to get people to go to screening, you need to publicise the success of other screening programmes, whether it's bowel, breast, cervical, whatever it is. Publicise how successful these are and really go for it and say, right, this is the next step in the screening and it's going to be a lung screening. (FG3, R1)\\nParticipants said that an invitation to screening should be encouraging without being coercive, should not mention the word ‘cancer’ too much and should not imply judgement about smoking behaviour:There's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\\n\\nThere's always going to be a worry anyway. But seeing the word ‘health check’ you would get more people to go. The word cancer and people just say, well I don't want to know. (FG2, R3)\",\n \"In addition to the implementation ideas given above, there was a very clear message from the focus groups when it came to addressing any access barriers to screening, relevant to both deprived and remote and rural communities—the use of mobile units. This, it was typically felt, would remove screening from a clinical environment, and bring it closer to local communities.So, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)This is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\n\\nSo, if there was screening available as a mobile unit or some sort of drop in people probably would be more willing to go and fit it in. It's just the same as these COVID jabs we've all had to do; if they were more available, a lot more people would do it. (FG2, R3)\\nThis is where the mobile vans come in, you know? They can drive to these remote areas, especially in Scotland, when they go further north. We've got two people from [place], […] some of the places are very remote so…. (FG1, R1)\\nWhile those in the rural‐dwelling focus group were unanimously in favour of mobile units, support for them was not confined to the rural group. Mobile units were supported as a way of engaging with marginalized groups, such as those who sought to avoid hospitals.\\nSupporting people who would lose pay if they attend a screening appointment was suggested as a statutory right, with employer support:I think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\n\\nI think your employer should get help if people need time off to go to these things, they should be encouraging and get paid for going to them, you know, get paid for so many hours, if you need so many hours off to go and get these tests done, you shouldn't lose your pay, because if you lose your pay, a lot of people will not go. (FG1, R3)\\nAnother suggestion for increasing participation was to ‘flood’ the public with information about screening through all media outlets about screening and engage with local communities, particularly the harder to reach groups, through ‘men's sheds’, workplaces, community hubs for older people as well as speaking to school pupils to normalize screening:And taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\\n\\nAnd taking the fear out of it by having all different age groups talking about it … It's trying to get the message out there, people need to start screening themselves from an age where they're invited to do so, and not be fearful of it. It's trying to get around that age group and succeeding with it, and hoping that future generations do go for their screenings when they're invited to do so. (FG2, R6)\",\n \"[SUBTITLE] Summary [SUBSECTION] This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\\nThis focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\\n[SUBTITLE] Comparison with existing literature and theory [SUBSECTION] There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\\n26\\n, \\n27\\n, \\n28\\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\\n27\\n, \\n29\\n, \\n30\\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\\n17\\n, \\n31\\n\\n\\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\\n28\\n, \\n32\\n, \\n33\\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\\n29\\n, \\n34\\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\\n35\\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\\n36\\n, \\n37\\n, \\n38\\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\\n28\\n, \\n33\\n, \\n39\\n, \\n40\\n Risk and decision‐making are discussed further below.\\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\\n31\\n, \\n38\\n, \\n41\\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\\n42\\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\\n17\\n, \\n29\\n, \\n38\\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\\n17\\n, \\n43\\n, \\n44\\n, \\n45\\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\\n46\\n, \\n47\\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\\n36\\n as well as being central to the long‐term cost‐effectiveness of screening.\\n13\\n\\n\\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\\n17\\n, \\n32\\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\\n48\\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\\n48\\n, \\n49\\n, \\n50\\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\\n51\\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\\n7\\n, \\n8\\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\\n52\\n, \\n53\\n\\n\\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\\n27\\n, \\n32\\n, \\n46\\n\\n\\nVon Wagner et al.\\n50\\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\\n54\\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\\n55\\n, \\n56\\n, \\n57\\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\\n37\\n, \\n39\\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\\n48\\n, \\n49\\n\\n\\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\\n58\\n, \\n59\\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\\n12\\n, \\n60\\n\\n\\nThere is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\\n26\\n, \\n27\\n, \\n28\\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\\n27\\n, \\n29\\n, \\n30\\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\\n17\\n, \\n31\\n\\n\\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\\n28\\n, \\n32\\n, \\n33\\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\\n29\\n, \\n34\\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\\n35\\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\\n36\\n, \\n37\\n, \\n38\\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\\n28\\n, \\n33\\n, \\n39\\n, \\n40\\n Risk and decision‐making are discussed further below.\\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\\n31\\n, \\n38\\n, \\n41\\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\\n42\\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\\n17\\n, \\n29\\n, \\n38\\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\\n17\\n, \\n43\\n, \\n44\\n, \\n45\\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\\n46\\n, \\n47\\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\\n36\\n as well as being central to the long‐term cost‐effectiveness of screening.\\n13\\n\\n\\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\\n17\\n, \\n32\\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\\n48\\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\\n48\\n, \\n49\\n, \\n50\\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\\n51\\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\\n7\\n, \\n8\\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\\n52\\n, \\n53\\n\\n\\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\\n27\\n, \\n32\\n, \\n46\\n\\n\\nVon Wagner et al.\\n50\\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\\n54\\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\\n55\\n, \\n56\\n, \\n57\\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\\n37\\n, \\n39\\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\\n48\\n, \\n49\\n\\n\\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\\n58\\n, \\n59\\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\\n12\\n, \\n60\\n\\n\\n[SUBTITLE] Strengths and limitations [SUBSECTION] This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\\n61\\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\\nThis set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\\n61\\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\\n[SUBTITLE] Implications and future work [SUBSECTION] This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\\n62\\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\\n40\\n, \\n45\\n, \\n46\\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\\n59\\n, \\n63\\n, \\n64\\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distress\\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\\nIncorporating discussion of fears associated with screening into information materials\\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\\nEnsuring a timely and sensitive approach to smoking cessation advice\\nThis study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\\n62\\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\\n40\\n, \\n45\\n, \\n46\\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\\n59\\n, \\n63\\n, \\n64\\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distress\\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\\nIncorporating discussion of fears associated with screening into information materials\\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\\nEnsuring a timely and sensitive approach to smoking cessation advice\",\n \"This focus group study is the first of its kind to ascertain the views of Scottish residents on a potential LDCT lung screening programme and identify likely barriers and facilitators in this context. Knowledge of lung screening was low among the focus group participants, although they showed awareness and personal experience of other screening programmes. Participants were very supportive of the idea of lung screening and harnessed the early detection narrative to discuss the importance of taking part. Participants reported the process of lung screening via a LDCT scan was acceptable. Two key barrier types were identified: individual level influences on screening intentions, and practical and system barriers. Within the individual level factors, emotional and psychological concerns related to fear of a cancer diagnosis, mistrust, fatalism and perceived stigma were dominant in focus group discussion. For some, screening was part of being a health conscious citizen and prioritizing health matters, while others based their decision‐making on their perceived risk. A number of potential practical barriers to lung screening participation were mooted that were particularly relevant for people living in deprived and rural communities. These included travel, cost, time and competing priorities. Maximizing accessibility was also key in the discussion and a distinct recommendation for future programmes.\",\n \"There is a strong consistency in our findings with the growing body of evidence looking at attitudes to screening and screening behaviour, across a range of screening programmes and for lung screening in particular, in the United Kingdom and beyond.\\n26\\n, \\n27\\n, \\n28\\n LDCT screening for lung cancer is largely unheard of in Scotland but has a high degree of acceptability more broadly, or among those who have participated in screening.\\n27\\n, \\n29\\n, \\n30\\n There was an evident awareness among participants of the benefits of early detection and thus support for lung screening, in line with other research.\\n17\\n, \\n31\\n\\n\\nFear was identified as one of the most common psychological barriers to lung screening, which is reflected in the literature in both survey and qualitative explorations of attitudes to screening.\\n28\\n, \\n32\\n, \\n33\\n Linked to this was a sense of fatalism or predicted fatalism among older generations, also mirrored in comparable studies of lung screening.\\n29\\n, \\n34\\n While there is evidence to suggest that the fear and anxiety associated with lung screening participation is transient\\n35\\n and can be a motivating factor to be screened, quit smoking and even other cancer‐preventing health behaviour change,\\n36\\n, \\n37\\n, \\n38\\n it is still vital to minimize this emotional response by addressing and managing it. Positive messaging in the promotion of screening, such as sharing the treatment successes and mortality gains from early detection, is one potential step in approaching this.\\nNotions of risk were a key component of our group's considerations of whether or not someone would take part in screening, often related to not experiencing symptoms and having stopped smoking. Perceived risk has been identified widely in the literature to explain decision‐making in relation to screening and help‐seeking for symptomatic illness.\\n28\\n, \\n33\\n, \\n39\\n, \\n40\\n Risk and decision‐making are discussed further below.\\nPerceived judgement and stigma related to smoking featured in focus group discussions. This is widely evident in the literature, along with self‐blame.\\n31\\n, \\n38\\n, \\n41\\n Stigma has been identified in the literature as a barrier to help‐seeking for signs of lung cancer and it seems this also applies to screening participation.\\n42\\n Related to this was a sense of fatalism—participants in our study did not consider they would blame themselves if they developed lung cancer, but some did feel that the damage was already done and screening could not change that.\\n17\\n, \\n29\\n, \\n38\\n However, this was not a clear barrier. For some, it was a good reason to detect any inevitable lung cancer at an early stage, suggesting that issues such as fear, blame and fatalism are complex and operate on a pendulum when prompting action or inaction.\\nDiscussions of stigma inevitably moved onto smoking cessation. Smoking cessation advice was broadly acceptable to our participants, but only if delivered in a noncoercive way; again, other studies have had similar findings.\\n17\\n, \\n43\\n, \\n44\\n, \\n45\\n In our focus group study, it appeared that participants may have become desensitized to smoking cessation advice. This has implications when considering brief interventions for smoking, and would need further exploration to understand whether small prompts may be enough to stimulate action among people who have intentions to quit smoking. There is evidence to suggest that brief interventions can be effective.\\n46\\n, \\n47\\n There is also emerging evidence that incorporating smoking cessation advice into lung screening is effective—seeing images of lungs has been a strong motivator to quit smoking\\n36\\n as well as being central to the long‐term cost‐effectiveness of screening.\\n13\\n\\n\\nSome participants in our focus groups described a difficult relationship and a level of mistrust in interacting with health professionals and services, often related to poor past experiences or a sociocultural divide. Such perceptions are often ingrained in more deprived communities and are often reported in the literature—relating not only to lung screening,\\n17\\n, \\n32\\n but beyond to studies of help‐seeking behaviour and doctor–patient relationships.\\n48\\n Again, this is a complex feature of health service engagement and can be linked to other barriers such as low self‐efficacy, low health literacy and power dynamics, which can be particularly divisive in terms of equity in access to health care for people in disadvantaged groups.\\n48\\n, \\n49\\n, \\n50\\n Development of interventions to repair broken relationships with the health service, a perceived authoritative institution, and other methods to improve accessibility such as communication training for professionals, targeted awareness campaigns and community engagement strategies can address inequities in access to screening services.\\n51\\n Primary care has an important role to play; current workforce shortages need to be addressed, and solutions identified which do not generate significant extra burden for primary care staff—one example is streamlining procedures to identify high‐risk patients from practice data.\\n7\\n, \\n8\\n It also seems logical to harness the successes of implementation strategies for other cancer screening (e.g., the UK's bowel screening programmes).\\n52\\n, \\n53\\n\\n\\nIn addition to the psychosocial issues discussed by focus group participants, practical barriers to participation in lung screening including time, cost, travel and competing work or other commitments, were also mooted. Practical barriers are commonly reported throughout the evidence base related to engaging with screening, with a suggestion that these barriers are heightened in more deprived groups.\\n27\\n, \\n32\\n, \\n46\\n\\n\\nVon Wagner et al.\\n50\\n have developed a model of screening behaviour, accounting for a range of factors identified in relation to wider health behaviours, such as health identity and self‐efficacy, that can be usefully mapped onto the findings of this study. Similarly, Robb\\n54\\n has developed the I‐SAM model to understand screening participation. Application of these models to follow‐up interviews as part of the planned pilot lung screening study in Scotland will be enlightening to confirm the salience of these to screening participation and nonparticipation.\\nThere is an abundance of early cancer detection research exploring how people appraise bodily changes, evaluate risk and decide to seek medical help, as well as conceptual models to understand these processes.\\n55\\n, \\n56\\n, \\n57\\n There is also some utility in applying these to screening behaviour, often in the absence of symptoms, to understand nonparticipation in screening. For example, Kummer et al.'s\\n37\\n, \\n39\\n cognitive heuristics for help‐seeking for cancer symptoms may act as prompts or inhibitors to participate in screening. Understanding these factors in the context of deprived populations adds further considerations that may compound behaviour in terms of available resources (cognitive, psychological and practical), competing demands and permeability of services.\\n48\\n, \\n49\\n\\n\\nIn terms of ideas to overcome barriers, participants focused primarily on positive messaging in information materials and advertising, and accessibility through the provision of mobile screening vans similar to those used in breast screening or for Covid‐19 vaccination. Positive and nonjudgemental messaging focusing on the gains on offer from screening and early detection has also been found elsewhere and incorporated into the pilot lung screening provision.\\n58\\n, \\n59\\n Travel to access screening services as well as fear of hospital environments have also been identified in the literature, and evidence evaluating the use of mobile screening vans to address these issues is beginning to accumulate.\\n12\\n, \\n60\\n\\n\",\n \"This set of focus groups provides a rich data set with an in‐depth discussion of the concept of lung cancer screening and anticipated barriers and facilitators. Focus group participants were a self‐selecting group who are accustomed to taking part in research studies, and who received a financial incentive. However, this form of recruitment strategy does allow access to people from a range of different backgrounds, including both deprived and rural areas of Scotland that may otherwise have been challenging to recruit. Three of the 24 participants were from ethnic minority groups, reflecting the general demographic in the Scottish population. Focus groups were conducted online. Video conferencing software facilitates easy interaction with people from around the country but can pose challenges. These include failing technology, building rapport and ensuring balanced participation within the group simply by being in a room together and acknowledging social cues. It was also necessary to consider that participants were talking about a hypothetical scenario and their intended behaviour, something we know does not simply translate into action.\\n61\\n Speaking to people who have taken part in screening or chosen not to take part is also essential to further understand the relationship between intention and behaviour.\\nWe also reflected on the group dynamic together with the nature of our role in conducting the focus groups, and whether this was likely to influence people in agreeing with and supporting the concept of lung screening. However, the open nature of questioning, reminding participants that we genuinely wanted to hear their views, and the self‐selecting group of individuals who were quite assured in their own responses, suggested that we did not shape this narrative.\",\n \"This study informs the development of strategies to improve uptake and informed choice in lung screening. It is essential to understand people's health beliefs and behaviours and to target the barriers to implement a patient‐centred service using a theory‐driven approach.\\n62\\n This work adds to a growing evidence base shedding light on the behavioural aspects of screening participation and will inform the design and implementation of a new lung screening pilot in Scotland\\n40\\n, \\n45\\n, \\n46\\n (see Box 2 for implementation ideas generated from this work). Minimizing practical barriers is also likely to be instrumental in improving participation and addressing inequity in access to screening. As such, information materials, methods of communication and the design of the process involved in screening (e.g., minimizing the steps, time commitment and waiting intervals), and sensitive, supportive messaging that addresses stigma and fear are all important components of a pilot to break down some of the known barriers. Drawing on the similarities with research based on other UK pilot studies, we are modelling optimized study and participant materials.\\n59\\n, \\n63\\n, \\n64\\n While there is consistency in the findings compared with existing work, it was important to explore the Scottish context with a diverse sample of participants to consider how rurality and deprivation presented any unique issues. Understanding ethnic variations in views on lung screening participation will also be important to ensure equitable provision and uptake of screening. A high burden of multimorbidity is also characteristic of the Scottish population and should be examined in future studies.\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distressAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐makingUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areasConsider covering the cost of travel expenses to screening facilities to ensure equitable accessIncorporating discussion of fears associated with screening into information materialsEnsuring positive messaging with nonjudgemental language around smoking behaviourEnsuring a timely and sensitive approach to smoking cessation advice\\n\\nMinimize steps in the screening process to lower opportunities for delays and associated distress\\nAvoiding unnecessary travel to scanning facilities with the chance to discuss screening concerns, address fears and perceived stigma, and facilitate informed decision‐making\\nUse of mobile screening vans as a ‘one stop shop’ to address resource constraints and travel issues for people living in deprived and rural areas\\nConsider covering the cost of travel expenses to screening facilities to ensure equitable access\\nIncorporating discussion of fears associated with screening into information materials\\nEnsuring positive messaging with nonjudgemental language around smoking behaviour\\nEnsuring a timely and sensitive approach to smoking cessation advice\",\n \"This focus group study has already identified several perceived individual, practical and system barriers and facilitators to participation in a pilot lung screening programme in Scotland using LDCT. While our results resonate with existing literature in this field, they will be helpful in addressing factors which are especially important in Scotland if it is to embrace lung cancer screening—reducing health inequalities, engaging deprived populations and ensuring access in remote and rural areas. The findings will inform the design and implementation of a Scottish pilot lung screening study.\",\n \"All authors apart from Jasmin Rostron and Lynsey R. Brown contributed toward the design of the focus groups and helped develop study documents. Debbie Cavers and Mia Nelson conducted the focus groups. Debbie Cavers, Mia Nelson and Jasmin Rostron coded and analysed the transcripts. Lynsey R. Brown read and commented on the analysis. Debbie Cavers drafted the manuscript and all authors edited and refined subsequent drafts.\",\n \"F. S. declares that the Universities of Dundee and St. Andrew's received funding from Oncimmune for the Early Detection of Cancer of the Lung Scotland (ECLS) trial from 2013 to 2021. The remaining authors declare no conflict of interest.\",\n \"The LUNGSCOT focus group study was approved by the University of Edinburgh Medical School Ethics Committee on 19 March 2021, reference 21‐EMREC‐002. All participants gave consent before taking part in the focus groups.\",\n \"Supporting information.\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,"discussion",null,null,null,null,"conclusions",null,"COI-statement",null,"supplementary-material"],"string":"[\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n null,\n \"conclusions\",\n null,\n \"COI-statement\",\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["early detection","focus group","lung cancer","lung screening","qualitative","screening"],"string":"[\n \"early detection\",\n \"focus group\",\n \"lung cancer\",\n \"lung screening\",\n \"qualitative\",\n \"screening\"\n]"}}},{"rowIdx":361,"cells":{"title":{"kind":"string","value":"The burden of chronic diseases and patients' preference for healthcare services among adult patients suffering from chronic diseases in Bangladesh."},"pmid":{"kind":"string","value":"36263949"},"background_abstract":{"kind":"string","value":"Low- and middle-income countries (LMICs) have a disproportionately high burden of chronic diseases, with inequalities in health care access and quality services. This study aimed to assess patients' preferences for healthcare services for chronic disease management among adult patients in Bangladesh."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"The present analysis was conducted among 10,385 patients suffering from chronic diseases, drawn from the latest Household Income and Expenditure Survey 2016-2017. We used the multinomial logistic regression to investigate the association of chronic comorbid conditions and healthcare service-related factors with patients' preferences for healthcare services."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The top four dimensions of patient preference for healthcare services in order of magnitude were quality of treatment (30.3%), short distance to health facility (27.6%), affordability of health care (21.7%) and availability of doctors (11.0%). Patients with heart disease had a 29% significantly lower preference for healthcare affordability than the quality of healthcare services (relative risk ratio [RRR] = 0.71; 0.56-0.90). Patients who received healthcare services from pharmacies or dispensaries were more likely to prefer a short distance to a health facility (RRR = 6.99; 4.80-9.86) or affordability of healthcare services (RRR = 3.13; 2.25-4.36). Patients with comorbid conditions were more likely to prefer healthcare affordability (RRR = 1.39; 1.15-1.68). In addition, patients who received health care from a public facility had 2.93 times higher preference for the availability of medical doctors (RRR = 2.93; 1.70-5.04) than the quality of treatment in the health facility, when compared with private service providers."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Patient preferences for healthcare services in chronic disease management were significantly associated with the type of disease and its magnitude and characteristics of healthcare providers. Therefore, to enhance service provision and equitable distribution and uptake of health services, policymakers and public health practitioners should consider patient preferences in designing national strategic frameworks for chronic disease management."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Adult","Humans","Patient Preference","Bangladesh","Health Services","Chronic Disease","Health Services Accessibility","Logistic Models"],"string":"[\n \"Adult\",\n \"Humans\",\n \"Patient Preference\",\n \"Bangladesh\",\n \"Health Services\",\n \"Chronic Disease\",\n \"Health Services Accessibility\",\n \"Logistic Models\"\n]"},"pmcid":{"kind":"string","value":"9700186"},"background_title":{"kind":"string","value":"BACKGROUND"},"background_text":{"kind":"string","value":"Chronic diseases have become a global challenge, imposing an enormous economic and health burden on society.\n1\n Chronic diseases are defined as health conditions lasting 12 months or more, which require ongoing medical intervention and may result in the limitation of activities of daily living.\n1\n The epidemiological burden of chronic diseases\n2\n and exposure to their risk factors are increasing worldwide,\n3\n particularly in low‐ and middle‐income countries (LMICs), such as Bangladesh.\n4\n, \n5\n, \n6\n Chronic diseases account for around 41 million deaths each year, representing about 71% of all deaths globally.\n7\n The most common chronic diseases include cardiovascular diseases (e.g., coronary heart diseases, stroke and peripheral vascular diseases), diabetes, cancers, chronic obstructive pulmonary disease, mental illness and arthritis\n7\n; approximately 77% of all yearly deaths are related to chronic diseases occur in LMICs, of which 85% occur in the most productive age groups (30–69 years).\n7\n It has been estimated that chronic diseases will account for an accumulated global economic loss of 47 trillion US dollars by 2030, approximately 75% of the global gross domestic product.\n8\n In Bangladesh, an LMIC with a substantial social and economic burden, about 886,000 deaths (i.e., 59% of total deaths) occur due to chronic diseases each year.\n3\n, \n9\n The burden due to chronic diseases has increased in Bangladesh from 43.4% in 2000 to 66.9% in 2015.\n3\n It is anticipated that chronic diseases will exceed the combined burden of communicable, maternal, perinatal and nutrition‐related diseases by 2030 globally, including in Bangladesh.\n10\n, \n11\n\n\nDespite the high burden of chronic diseases, Bangladesh, like many LMIC countries, does not have a national integrated chronic diseases management policy, strategy or action plan.\n9\n, \n11\n For instance, the prevalence of undiagnosed chronic diseases is high, and the proportion of unmanaged chronic diseases is even higher in many LMICs,\n10\n including Bangladesh.\n12\n This highlights the frequent inadequacies in the diagnosis, prevention and management of chronic diseases among the healthcare systems of LMICs.\n9\n, \n10\n Efforts in chronic disease management in Bangladesh continue to be inadequate. Little attention has been given to addressing the contributing behaviours associated with chronic diseases, including unhealthy dietary patterns, lack of physical activity and exposure to factors potentially detrimental to health such as alcohol, drug and tobacco use.\n6\n However, it is possible to counteract the rising prevalence of the chronic disease by implementing effective prevention strategies, population‐based screening, reduction of risk factors, early detection and appropriate treatments.\n10\n, \n11\n If such actions are not taken, the burden of chronic diseases, which is referred to as an emerging prevalence of chronic diseases globally imposing an enormous economic and health burden will likely continue to rise,\n10\n, \n11\n which is alarming especially among the vulnerable and marginalized populations of Bangladesh with limited affordability for health services.\n11\n In addition, the emerging prevalence of chronic disease may also lead to a health system burden in terms of increasing healthcare utilization, treatment costs and chronic disease management.\nIt is well established that chronic diseases are increasingly associated with the over‐utilization of healthcare services and a higher financial burden.\n13\n For example, a previous study documented that a higher number of chronic diseases was linked to an increased number of outpatient visits.\n14\n Therefore, adequate preventative services must be in place to reduce the social‐economic burden of chronic disease, thereby ensuring optimal use of health resources. However, the major challenges to ensure effective prevention and management of chronic diseases include social status, power gradients, racial/ethnic differences, poor accessibility and affordability of healthcare services.\n2\n, \n6\n, \n15\n, \n16\n, \n17\n Previous research has identified several factors associated with healthcare utilization, such as demographic and socioeconomic characteristics, type of healthcare providers and presence of chronic illnesses.\n18\n In addition, patients' preference for healthcare services depends on several factors, including personal preference, disease severity, economic capacity, the reputation of healthcare providers\n18\n as well as affordable costs associated with treatment.\n19\n Furthermore, short travel time to healthcare facilities, effective interactions with healthcare providers,\n19\n, \n20\n respectful service provider attitudes\n21\n and short waiting time\n22\n were positively associated with the patient preference for healthcare services. A recent discrete choice experiment study found that the availability of medicine and transport to the health facility were significant attributes of patient preference for healthcare services.\n23\n Notably, optimum healthcare utilization among chronically ill patients has a significant role in preventing and managing chronic diseases.\nIn Bangladesh, a family spends an average of 11% of their total household budget on health care and half of the population spends 7% of their monthly per capita consumption expenditure on illness.\n24\n Understanding patients' preferences could help medical professionals and healthcare providers restructure the healthcare delivery model and ensure the quality of services. In addition to clinical guidelines, patients' preferences may also provide guidelines for the selection of treatment options. Patient preferences also help to inform clinical decisions where science has not yet been able to provide effective solutions to healthcare problems.\n25\n Therefore, information on patients' preferences for healthcare utilization in terms of healthcare expenditure is critical; without this knowledge, an effective national healthcare policy cannot be formulated.\n26\n Moreover, there is a current lack of evidence on the patients' preferences for healthcare services in chronic disease management in Bangladesh. Since no such study exists, the present study aimed to investigate patients' preference for healthcare services among adult patients in Bangladesh. Understanding patients' preferences are important in designing interventions aimed at reducing the burden of chronic diseases, which undermine economic growth and development and are increasingly becoming the leading cause of death."},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Study design and data sources [SUBSECTION] This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\nThis study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\n[SUBTITLE] HIES survey sampling and sample size calculation [SUBSECTION] The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\nThe sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"Table 1 shows the participant's characteristics. Of the 10,385 patients with chronic diseases, approximately 50% were female, 51% were married and 70% were unemployed. The majority of the patients were aged between 18 and 45 (81%) years, and around one‐third (37%) had no formal education.\nDistribution of participant's characteristics\n\nNote: p value was derived using χ\n2 test or one‐way analysis of variance where appropriate.\nAbbreviations: BDT, Bangladeshi Taka; CI, confidence interval; SD, standard deviation.\n[SUBTITLE] The distribution of chronic disease and utilization of healthcare services [SUBSECTION] The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\nThe most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\n[SUBTITLE] Preferences for healthcare services for chronic illnesses [SUBSECTION] Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\nApproximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\n[SUBTITLE] Correlates of patient preference for chronic disease‐related healthcare services [SUBSECTION] Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\nTable 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors."},"conclusions_title":{"kind":"string","value":"CONCLUSION"},"conclusions_text":{"kind":"string","value":"Our study findings highlighted that patient preferences for healthcare services in chronic disease management were significantly associated with disease severity and healthcare providers' attributes. Therefore, policymakers and public health practitioners should consider patient preferences for managing chronic conditions within the national strategic frameworks to improve service provision, equitable distribution and uptake of the services."},"other_sections_titles":{"kind":"list like","value":["Study design and data sources","HIES survey sampling and sample size calculation","DATA COLLECTION","Study population","DEFINITIONS OF STUDY VARIABLES","Outcome measures","Chronic diseases and comorbid conditions","Covariates","STATISTICAL ANALYSIS","The distribution of chronic disease and utilization of healthcare services","Preferences for healthcare services for chronic illnesses","Correlates of patient preference for chronic disease‐related healthcare services","Implications for policy and practice","AUTHOR CONTRIBUTIONS","ETHICS STATEMENT"],"string":"[\n \"Study design and data sources\",\n \"HIES survey sampling and sample size calculation\",\n \"DATA COLLECTION\",\n \"Study population\",\n \"DEFINITIONS OF STUDY VARIABLES\",\n \"Outcome measures\",\n \"Chronic diseases and comorbid conditions\",\n \"Covariates\",\n \"STATISTICAL ANALYSIS\",\n \"The distribution of chronic disease and utilization of healthcare services\",\n \"Preferences for healthcare services for chronic illnesses\",\n \"Correlates of patient preference for chronic disease‐related healthcare services\",\n \"Implications for policy and practice\",\n \"AUTHOR CONTRIBUTIONS\",\n \"ETHICS STATEMENT\"\n]"},"other_sections_texts":{"kind":"list like","value":["This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.","The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.","Data collection was accomplished between early April 2016 and late March 2017 through face‐to‐face interviews with the participants. A total of 128 enumeration teams, each comprised of five members (one supervising officer, two interviewers and two female facilitators) who were thoroughly trained, completed the data collection. A multilayered quality control measure was employed to ensure the quality of the collected data. The supervising officer of each team was responsible to verify all the questionnaires completed by the field staff before they were sent to the headquarter. Moreover, Deputy Directors of the District Statistical Offices as well as senior officials from the headquarter, frequently visited the sampled areas to ensure data quality (Supporting Information: Document 1).\nThe survey team formulated an operational definition of each variable used in the questionnaire. For example, while collecting chronic disease‐related information, the enumerators were instructed to ask the respondents, ‘Have you suffered from any chronic illness/disability in the last 12 months or more?’ (if yes); then, participants were asked a second question: ‘What chronic illness/disability are you suffering from?’ with response options: (1) chronic fever, (2) Injuries/disability, (3) chronic heart disease, (4) respiratory diseases/asthma/bronchitis, (5) diarrhoea/dysentery, (6) gastric/ulcer, (7) blood pressure, (8) arthritis/rheumatism, (9) skin problem, (10) diabetes, (11) cancer, (12) kidney diseases, (13) liver diseases, (14) mental health, (15) paralysis, (16) ear/ENT problem, (17) eye problem or (18) other (specify) (Supporting Information: Appendix Table A1: HIES 2016–2017 questionnaire). Participants responded based on their disease diagnosis, experiences, symptoms of illness and course of treatment. To get valid information, the enumerators also probed where necessary, asked the respondents to show any relevant documents such as prescriptions and test reports, or explained to the respondents about chronic diseases using various case scenarios as outlined in the survey guidelines and instructions.\n[SUBTITLE] Study population [SUBSECTION] All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants\nAll selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants","All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants","[SUBTITLE] Outcome measures [SUBSECTION] Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\nPatients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\n[SUBTITLE] Chronic diseases and comorbid conditions [SUBSECTION] Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\nParticipants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\n[SUBTITLE] Covariates [SUBSECTION] Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\nBased on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).","Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).","Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.","Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).","The adjusted multinomial logistic regression model was used to identify the potential factors that had an association with the patient's preference for healthcare service. The dependent variable (chronic illness patient's preference for healthcare services) was characterized as a categorical measure in the regression model. An unadjusted analysis was performed using each of the explanatory variables for the following reasons: (1) primary screening of the selection of qualified variables, which were added in the adjusted model, (2) the χ\n2 tests (or one‐way analysis where appropriate) were used to find the association between outcome and explanatory variables. However, the majority of the explanatory variables were categorical with two or more labels; therefore, an unadjusted analysis was performed to find the association between the outcome variable and different categories of explanatory variables. The explanatory variables were included in the adjusted model only if any label of the predictor was significant at a 5% or less risk level in the unadjusted model, which in turn was used to adjust for the associations of other potential confounders. For the explanatory variables, the category found to be least at risk of having patients' preferences for healthcare services related to chronic illness in the analysis was considered the reference category for constructing the relative risk ratio (RRR). Statistical significance was considered at a 5% risk level. All data analyses were undertaken using the statistical software Stata/SE 14 (StataCorp).","The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).","Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.","Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.","Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\n48\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\n9\n, \n11\n, \n49\n\n\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\n11\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\n48\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\n11\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.","Rashidul A. Mahumud conceptualized the study, directed the data analysis and wrote most of the manuscript. Rashidul A. Mahumud performed the statistical analysis under the supervision of Andre M. N. Renzaho, Khorshed Alam and Asaduzzaman Khan. Marufa Sultana, Satyajit Kundu, Md. A. Rahman and Mostafa Kamal wrote portions of the manuscript. Rashidul A. Mahumud, Khorshed Alam, Andre M. N. Renzaho, Sabuj K. Mistry, Joseph K. Kamara, Md. G. Hossain, Mohammad A. Ali and Asaduzzaman Khan contributed to the study design, data interpretation and edited the manuscript. All named authors contributed to critically reviewing the revised initial draft of the manuscript. All authors read and approved the final draft.","This study has been prepared using a secondary de‐identified and fully anonymized cross‐sectional data set from the HIES conducted during 2016–2017 by the Bangladesh Bureau of Statistics. Therefore, no ethical approval was not required for this study."],"string":"[\n \"This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\\n27\\n\\n\\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\",\n \"The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\\n27\\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\\n27\\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\\n27\\n using the following formula\\n\\n(1)\\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\",\n \"Data collection was accomplished between early April 2016 and late March 2017 through face‐to‐face interviews with the participants. A total of 128 enumeration teams, each comprised of five members (one supervising officer, two interviewers and two female facilitators) who were thoroughly trained, completed the data collection. A multilayered quality control measure was employed to ensure the quality of the collected data. The supervising officer of each team was responsible to verify all the questionnaires completed by the field staff before they were sent to the headquarter. Moreover, Deputy Directors of the District Statistical Offices as well as senior officials from the headquarter, frequently visited the sampled areas to ensure data quality (Supporting Information: Document 1).\\nThe survey team formulated an operational definition of each variable used in the questionnaire. For example, while collecting chronic disease‐related information, the enumerators were instructed to ask the respondents, ‘Have you suffered from any chronic illness/disability in the last 12 months or more?’ (if yes); then, participants were asked a second question: ‘What chronic illness/disability are you suffering from?’ with response options: (1) chronic fever, (2) Injuries/disability, (3) chronic heart disease, (4) respiratory diseases/asthma/bronchitis, (5) diarrhoea/dysentery, (6) gastric/ulcer, (7) blood pressure, (8) arthritis/rheumatism, (9) skin problem, (10) diabetes, (11) cancer, (12) kidney diseases, (13) liver diseases, (14) mental health, (15) paralysis, (16) ear/ENT problem, (17) eye problem or (18) other (specify) (Supporting Information: Appendix Table A1: HIES 2016–2017 questionnaire). Participants responded based on their disease diagnosis, experiences, symptoms of illness and course of treatment. To get valid information, the enumerators also probed where necessary, asked the respondents to show any relevant documents such as prescriptions and test reports, or explained to the respondents about chronic diseases using various case scenarios as outlined in the survey guidelines and instructions.\\n[SUBTITLE] Study population [SUBSECTION] All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\\nAll selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\",\n \"All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\",\n \"[SUBTITLE] Outcome measures [SUBSECTION] Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\\nPatients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\\n[SUBTITLE] Chronic diseases and comorbid conditions [SUBSECTION] Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\\nParticipants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\\n[SUBTITLE] Covariates [SUBSECTION] Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\\nBased on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\",\n \"Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\",\n \"Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\",\n \"Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\",\n \"The adjusted multinomial logistic regression model was used to identify the potential factors that had an association with the patient's preference for healthcare service. The dependent variable (chronic illness patient's preference for healthcare services) was characterized as a categorical measure in the regression model. An unadjusted analysis was performed using each of the explanatory variables for the following reasons: (1) primary screening of the selection of qualified variables, which were added in the adjusted model, (2) the χ\\n2 tests (or one‐way analysis where appropriate) were used to find the association between outcome and explanatory variables. However, the majority of the explanatory variables were categorical with two or more labels; therefore, an unadjusted analysis was performed to find the association between the outcome variable and different categories of explanatory variables. The explanatory variables were included in the adjusted model only if any label of the predictor was significant at a 5% or less risk level in the unadjusted model, which in turn was used to adjust for the associations of other potential confounders. For the explanatory variables, the category found to be least at risk of having patients' preferences for healthcare services related to chronic illness in the analysis was considered the reference category for constructing the relative risk ratio (RRR). Statistical significance was considered at a 5% risk level. All data analyses were undertaken using the statistical software Stata/SE 14 (StataCorp).\",\n \"The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\",\n \"Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\",\n \"Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\\nInfluencing factors on patient's preference of healthcare services\\n\\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\",\n \"Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\\n48\\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\\n9\\n, \\n11\\n, \\n49\\n\\n\\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\\n11\\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\\n48\\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\\n11\\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.\",\n \"Rashidul A. Mahumud conceptualized the study, directed the data analysis and wrote most of the manuscript. Rashidul A. Mahumud performed the statistical analysis under the supervision of Andre M. N. Renzaho, Khorshed Alam and Asaduzzaman Khan. Marufa Sultana, Satyajit Kundu, Md. A. Rahman and Mostafa Kamal wrote portions of the manuscript. Rashidul A. Mahumud, Khorshed Alam, Andre M. N. Renzaho, Sabuj K. Mistry, Joseph K. Kamara, Md. G. Hossain, Mohammad A. Ali and Asaduzzaman Khan contributed to the study design, data interpretation and edited the manuscript. All named authors contributed to critically reviewing the revised initial draft of the manuscript. All authors read and approved the final draft.\",\n \"This study has been prepared using a secondary de‐identified and fully anonymized cross‐sectional data set from the HIES conducted during 2016–2017 by the Bangladesh Bureau of Statistics. Therefore, no ethical approval was not required for this study.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["BACKGROUND","METHODS","Study design and data sources","HIES survey sampling and sample size calculation","DATA COLLECTION","Study population","DEFINITIONS OF STUDY VARIABLES","Outcome measures","Chronic diseases and comorbid conditions","Covariates","STATISTICAL ANALYSIS","RESULTS","The distribution of chronic disease and utilization of healthcare services","Preferences for healthcare services for chronic illnesses","Correlates of patient preference for chronic disease‐related healthcare services","DISCUSSION","Implications for policy and practice","CONCLUSION","AUTHOR CONTRIBUTIONS","CONFLICT OF INTEREST","ETHICS STATEMENT","Supporting information"],"string":"[\n \"BACKGROUND\",\n \"METHODS\",\n \"Study design and data sources\",\n \"HIES survey sampling and sample size calculation\",\n \"DATA COLLECTION\",\n \"Study population\",\n \"DEFINITIONS OF STUDY VARIABLES\",\n \"Outcome measures\",\n \"Chronic diseases and comorbid conditions\",\n \"Covariates\",\n \"STATISTICAL ANALYSIS\",\n \"RESULTS\",\n \"The distribution of chronic disease and utilization of healthcare services\",\n \"Preferences for healthcare services for chronic illnesses\",\n \"Correlates of patient preference for chronic disease‐related healthcare services\",\n \"DISCUSSION\",\n \"Implications for policy and practice\",\n \"CONCLUSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"CONFLICT OF INTEREST\",\n \"ETHICS STATEMENT\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Chronic diseases have become a global challenge, imposing an enormous economic and health burden on society.\n1\n Chronic diseases are defined as health conditions lasting 12 months or more, which require ongoing medical intervention and may result in the limitation of activities of daily living.\n1\n The epidemiological burden of chronic diseases\n2\n and exposure to their risk factors are increasing worldwide,\n3\n particularly in low‐ and middle‐income countries (LMICs), such as Bangladesh.\n4\n, \n5\n, \n6\n Chronic diseases account for around 41 million deaths each year, representing about 71% of all deaths globally.\n7\n The most common chronic diseases include cardiovascular diseases (e.g., coronary heart diseases, stroke and peripheral vascular diseases), diabetes, cancers, chronic obstructive pulmonary disease, mental illness and arthritis\n7\n; approximately 77% of all yearly deaths are related to chronic diseases occur in LMICs, of which 85% occur in the most productive age groups (30–69 years).\n7\n It has been estimated that chronic diseases will account for an accumulated global economic loss of 47 trillion US dollars by 2030, approximately 75% of the global gross domestic product.\n8\n In Bangladesh, an LMIC with a substantial social and economic burden, about 886,000 deaths (i.e., 59% of total deaths) occur due to chronic diseases each year.\n3\n, \n9\n The burden due to chronic diseases has increased in Bangladesh from 43.4% in 2000 to 66.9% in 2015.\n3\n It is anticipated that chronic diseases will exceed the combined burden of communicable, maternal, perinatal and nutrition‐related diseases by 2030 globally, including in Bangladesh.\n10\n, \n11\n\n\nDespite the high burden of chronic diseases, Bangladesh, like many LMIC countries, does not have a national integrated chronic diseases management policy, strategy or action plan.\n9\n, \n11\n For instance, the prevalence of undiagnosed chronic diseases is high, and the proportion of unmanaged chronic diseases is even higher in many LMICs,\n10\n including Bangladesh.\n12\n This highlights the frequent inadequacies in the diagnosis, prevention and management of chronic diseases among the healthcare systems of LMICs.\n9\n, \n10\n Efforts in chronic disease management in Bangladesh continue to be inadequate. Little attention has been given to addressing the contributing behaviours associated with chronic diseases, including unhealthy dietary patterns, lack of physical activity and exposure to factors potentially detrimental to health such as alcohol, drug and tobacco use.\n6\n However, it is possible to counteract the rising prevalence of the chronic disease by implementing effective prevention strategies, population‐based screening, reduction of risk factors, early detection and appropriate treatments.\n10\n, \n11\n If such actions are not taken, the burden of chronic diseases, which is referred to as an emerging prevalence of chronic diseases globally imposing an enormous economic and health burden will likely continue to rise,\n10\n, \n11\n which is alarming especially among the vulnerable and marginalized populations of Bangladesh with limited affordability for health services.\n11\n In addition, the emerging prevalence of chronic disease may also lead to a health system burden in terms of increasing healthcare utilization, treatment costs and chronic disease management.\nIt is well established that chronic diseases are increasingly associated with the over‐utilization of healthcare services and a higher financial burden.\n13\n For example, a previous study documented that a higher number of chronic diseases was linked to an increased number of outpatient visits.\n14\n Therefore, adequate preventative services must be in place to reduce the social‐economic burden of chronic disease, thereby ensuring optimal use of health resources. However, the major challenges to ensure effective prevention and management of chronic diseases include social status, power gradients, racial/ethnic differences, poor accessibility and affordability of healthcare services.\n2\n, \n6\n, \n15\n, \n16\n, \n17\n Previous research has identified several factors associated with healthcare utilization, such as demographic and socioeconomic characteristics, type of healthcare providers and presence of chronic illnesses.\n18\n In addition, patients' preference for healthcare services depends on several factors, including personal preference, disease severity, economic capacity, the reputation of healthcare providers\n18\n as well as affordable costs associated with treatment.\n19\n Furthermore, short travel time to healthcare facilities, effective interactions with healthcare providers,\n19\n, \n20\n respectful service provider attitudes\n21\n and short waiting time\n22\n were positively associated with the patient preference for healthcare services. A recent discrete choice experiment study found that the availability of medicine and transport to the health facility were significant attributes of patient preference for healthcare services.\n23\n Notably, optimum healthcare utilization among chronically ill patients has a significant role in preventing and managing chronic diseases.\nIn Bangladesh, a family spends an average of 11% of their total household budget on health care and half of the population spends 7% of their monthly per capita consumption expenditure on illness.\n24\n Understanding patients' preferences could help medical professionals and healthcare providers restructure the healthcare delivery model and ensure the quality of services. In addition to clinical guidelines, patients' preferences may also provide guidelines for the selection of treatment options. Patient preferences also help to inform clinical decisions where science has not yet been able to provide effective solutions to healthcare problems.\n25\n Therefore, information on patients' preferences for healthcare utilization in terms of healthcare expenditure is critical; without this knowledge, an effective national healthcare policy cannot be formulated.\n26\n Moreover, there is a current lack of evidence on the patients' preferences for healthcare services in chronic disease management in Bangladesh. Since no such study exists, the present study aimed to investigate patients' preference for healthcare services among adult patients in Bangladesh. Understanding patients' preferences are important in designing interventions aimed at reducing the burden of chronic diseases, which undermine economic growth and development and are increasingly becoming the leading cause of death.","[SUBTITLE] Study design and data sources [SUBSECTION] This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\nThis study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\n[SUBTITLE] HIES survey sampling and sample size calculation [SUBSECTION] The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\nThe sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.","This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\n27\n\n\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.","The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\n27\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\n27\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\n27\n using the following formula\n\n(1)\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.","Data collection was accomplished between early April 2016 and late March 2017 through face‐to‐face interviews with the participants. A total of 128 enumeration teams, each comprised of five members (one supervising officer, two interviewers and two female facilitators) who were thoroughly trained, completed the data collection. A multilayered quality control measure was employed to ensure the quality of the collected data. The supervising officer of each team was responsible to verify all the questionnaires completed by the field staff before they were sent to the headquarter. Moreover, Deputy Directors of the District Statistical Offices as well as senior officials from the headquarter, frequently visited the sampled areas to ensure data quality (Supporting Information: Document 1).\nThe survey team formulated an operational definition of each variable used in the questionnaire. For example, while collecting chronic disease‐related information, the enumerators were instructed to ask the respondents, ‘Have you suffered from any chronic illness/disability in the last 12 months or more?’ (if yes); then, participants were asked a second question: ‘What chronic illness/disability are you suffering from?’ with response options: (1) chronic fever, (2) Injuries/disability, (3) chronic heart disease, (4) respiratory diseases/asthma/bronchitis, (5) diarrhoea/dysentery, (6) gastric/ulcer, (7) blood pressure, (8) arthritis/rheumatism, (9) skin problem, (10) diabetes, (11) cancer, (12) kidney diseases, (13) liver diseases, (14) mental health, (15) paralysis, (16) ear/ENT problem, (17) eye problem or (18) other (specify) (Supporting Information: Appendix Table A1: HIES 2016–2017 questionnaire). Participants responded based on their disease diagnosis, experiences, symptoms of illness and course of treatment. To get valid information, the enumerators also probed where necessary, asked the respondents to show any relevant documents such as prescriptions and test reports, or explained to the respondents about chronic diseases using various case scenarios as outlined in the survey guidelines and instructions.\n[SUBTITLE] Study population [SUBSECTION] All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants\nAll selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants","All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\nDistribution of study participants","[SUBTITLE] Outcome measures [SUBSECTION] Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\nPatients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\n[SUBTITLE] Chronic diseases and comorbid conditions [SUBSECTION] Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\nParticipants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\n[SUBTITLE] Covariates [SUBSECTION] Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\nBased on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).","Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).","Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\n28\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\n28\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\n29\n, \n30\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.","Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).","The adjusted multinomial logistic regression model was used to identify the potential factors that had an association with the patient's preference for healthcare service. The dependent variable (chronic illness patient's preference for healthcare services) was characterized as a categorical measure in the regression model. An unadjusted analysis was performed using each of the explanatory variables for the following reasons: (1) primary screening of the selection of qualified variables, which were added in the adjusted model, (2) the χ\n2 tests (or one‐way analysis where appropriate) were used to find the association between outcome and explanatory variables. However, the majority of the explanatory variables were categorical with two or more labels; therefore, an unadjusted analysis was performed to find the association between the outcome variable and different categories of explanatory variables. The explanatory variables were included in the adjusted model only if any label of the predictor was significant at a 5% or less risk level in the unadjusted model, which in turn was used to adjust for the associations of other potential confounders. For the explanatory variables, the category found to be least at risk of having patients' preferences for healthcare services related to chronic illness in the analysis was considered the reference category for constructing the relative risk ratio (RRR). Statistical significance was considered at a 5% risk level. All data analyses were undertaken using the statistical software Stata/SE 14 (StataCorp).","Table 1 shows the participant's characteristics. Of the 10,385 patients with chronic diseases, approximately 50% were female, 51% were married and 70% were unemployed. The majority of the patients were aged between 18 and 45 (81%) years, and around one‐third (37%) had no formal education.\nDistribution of participant's characteristics\n\nNote: p value was derived using χ\n2 test or one‐way analysis of variance where appropriate.\nAbbreviations: BDT, Bangladeshi Taka; CI, confidence interval; SD, standard deviation.\n[SUBTITLE] The distribution of chronic disease and utilization of healthcare services [SUBSECTION] The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\nThe most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\n[SUBTITLE] Preferences for healthcare services for chronic illnesses [SUBSECTION] Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\nApproximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\n[SUBTITLE] Correlates of patient preference for chronic disease‐related healthcare services [SUBSECTION] Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\nTable 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.","The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).","Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.","Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\nInfluencing factors on patient's preference of healthcare services\n\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.","Chronic diseases among adults are becoming a significant health concern in many LMICs, including Bangladesh. This is the first study to focus on patient preferences for healthcare services for chronic disease management, using a recent nationally representative HIES. This study provided evidence of patient preference for the utilization of healthcare services in chronic disease management. The top four dimensions of patient preference for health care in order of preference were quality of treatment (30.3%), short distance to health facilities (27.6%), affordability of health care (21.7%) and availability of doctors (11.0%). The severity of the disease and the characteristics of healthcare providers were the most important contributing factors in patients' preferences and decisions to seek healthcare services.\nThis study showed that patients with heart disease were more likely to prefer quality health care than healthcare affordability or a short distance to a health facility. This is in agreement with previous studies, which have also documented that chronic heart disease patients are more likely to prefer quality healthcare services.\n31\n, \n32\n, \n33\n Several attributes might influence a patient's preference for quality healthcare services. For instance, heart disease may be associated with both acute episodes and high levels of long‐term adverse events (e.g., mortality and disability).\n34\n, \n35\n, \n36\n Such patients may require curative care as a component of a continuous, coordinated care model, covering both primary and postacute hospital care.\n34\n Moreover, due to the high disease burden and prolonged treatment, continuous use of healthcare resources (e.g., specialist consultation, diagnostic and medicine) may lead to a higher economic burden for households, individuals and society.\n35\n, \n36\n, \n37\n From a healthcare service providers' perspective, examining patients' experience and satisfaction with healthcare services could identify whether services are of an acceptable standard or highlight areas for potential quality improvement.\n38\n Existing research has revealed that a positive experience with healthcare professionals or other medical staff is directly related to patient preference for healthcare quality during hospitalization or course of treatment.\n39\n Other studies have indicated the importance of having confidence in the expertise and attentiveness of doctors or nurses,\n40\n, \n41\n healthcare provider's interpersonal communication skills and behaviours,\n42\n although shorter waiting times have also been mentioned/raised.\n43\n\n\nIn the present study, one of the most influential aspects of a patient's healthcare‐seeking behaviour was the healthcare provider's location; the shorter the distance to the healthcare facility the more likely the patient uptake of the healthcare services. It is plausible that patients may be unwilling or unable to travel long distances to access medical expertise or treatment, particularly if the nature of the chronic illness requires frequent appointments. This finding is supported by previous studies which reported that distance to healthcare facilities was a potential barrier to accessing healthcare services in LMICs.\n44\n, \n45\n Patients' educational level, employment status and chronic comorbid conditions were also identified as significant determinants influencing their healthcare‐seeking preferences. This may be due to patients with chronic illnesses being more likely to prioritize treatment, management and quality of life over other life choices they may face.\n46\n In addition, patients with multiple chronic conditions may require medical care from several healthcare providers across various locations throughout the year. This may be attributed to seeking quality care from experienced or specialist medical professionals, and superior diagnosis and treatment facilities, which has also been identified elsewhere.\n47\n\n\nAlthough this is the first study to investigate patient preference for healthcare services to manage chronic diseases among adults in Bangladesh, the extension and transferability of our findings to contexts beyond the study population should be handled with caution because of our study limitations. For instance, due to the cross‐sectional nature of the study, causality cannot be inferred. Furthermore, these findings may be subjected to some level of bias as data on the main variables of interest were self‐reported (i.e., illness, utilization and expenditure), thereby risking recall bias. However, the relatively short recall period (i.e., the last 30 days) of the household income and expenditure survey strives to reduce this potential bias.\n[SUBTITLE] Implications for policy and practice [SUBSECTION] Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\n48\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\n9\n, \n11\n, \n49\n\n\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\n11\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\n48\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\n11\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.\nOur study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\n48\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\n9\n, \n11\n, \n49\n\n\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\n11\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\n48\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\n11\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.","Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\n48\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\n9\n, \n11\n, \n49\n\n\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\n11\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\n48\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\n11\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.","Our study findings highlighted that patient preferences for healthcare services in chronic disease management were significantly associated with disease severity and healthcare providers' attributes. Therefore, policymakers and public health practitioners should consider patient preferences for managing chronic conditions within the national strategic frameworks to improve service provision, equitable distribution and uptake of the services.","Rashidul A. Mahumud conceptualized the study, directed the data analysis and wrote most of the manuscript. Rashidul A. Mahumud performed the statistical analysis under the supervision of Andre M. N. Renzaho, Khorshed Alam and Asaduzzaman Khan. Marufa Sultana, Satyajit Kundu, Md. A. Rahman and Mostafa Kamal wrote portions of the manuscript. Rashidul A. Mahumud, Khorshed Alam, Andre M. N. Renzaho, Sabuj K. Mistry, Joseph K. Kamara, Md. G. Hossain, Mohammad A. Ali and Asaduzzaman Khan contributed to the study design, data interpretation and edited the manuscript. All named authors contributed to critically reviewing the revised initial draft of the manuscript. All authors read and approved the final draft.","The authors declare no conflict of interest.","This study has been prepared using a secondary de‐identified and fully anonymized cross‐sectional data set from the HIES conducted during 2016–2017 by the Bangladesh Bureau of Statistics. Therefore, no ethical approval was not required for this study.","Supporting information.\nClick here for additional data file.\nSupporting information.\nClick here for additional data file.\nSupporting information.\nClick here for additional data file."],"string":"[\n \"Chronic diseases have become a global challenge, imposing an enormous economic and health burden on society.\\n1\\n Chronic diseases are defined as health conditions lasting 12 months or more, which require ongoing medical intervention and may result in the limitation of activities of daily living.\\n1\\n The epidemiological burden of chronic diseases\\n2\\n and exposure to their risk factors are increasing worldwide,\\n3\\n particularly in low‐ and middle‐income countries (LMICs), such as Bangladesh.\\n4\\n, \\n5\\n, \\n6\\n Chronic diseases account for around 41 million deaths each year, representing about 71% of all deaths globally.\\n7\\n The most common chronic diseases include cardiovascular diseases (e.g., coronary heart diseases, stroke and peripheral vascular diseases), diabetes, cancers, chronic obstructive pulmonary disease, mental illness and arthritis\\n7\\n; approximately 77% of all yearly deaths are related to chronic diseases occur in LMICs, of which 85% occur in the most productive age groups (30–69 years).\\n7\\n It has been estimated that chronic diseases will account for an accumulated global economic loss of 47 trillion US dollars by 2030, approximately 75% of the global gross domestic product.\\n8\\n In Bangladesh, an LMIC with a substantial social and economic burden, about 886,000 deaths (i.e., 59% of total deaths) occur due to chronic diseases each year.\\n3\\n, \\n9\\n The burden due to chronic diseases has increased in Bangladesh from 43.4% in 2000 to 66.9% in 2015.\\n3\\n It is anticipated that chronic diseases will exceed the combined burden of communicable, maternal, perinatal and nutrition‐related diseases by 2030 globally, including in Bangladesh.\\n10\\n, \\n11\\n\\n\\nDespite the high burden of chronic diseases, Bangladesh, like many LMIC countries, does not have a national integrated chronic diseases management policy, strategy or action plan.\\n9\\n, \\n11\\n For instance, the prevalence of undiagnosed chronic diseases is high, and the proportion of unmanaged chronic diseases is even higher in many LMICs,\\n10\\n including Bangladesh.\\n12\\n This highlights the frequent inadequacies in the diagnosis, prevention and management of chronic diseases among the healthcare systems of LMICs.\\n9\\n, \\n10\\n Efforts in chronic disease management in Bangladesh continue to be inadequate. Little attention has been given to addressing the contributing behaviours associated with chronic diseases, including unhealthy dietary patterns, lack of physical activity and exposure to factors potentially detrimental to health such as alcohol, drug and tobacco use.\\n6\\n However, it is possible to counteract the rising prevalence of the chronic disease by implementing effective prevention strategies, population‐based screening, reduction of risk factors, early detection and appropriate treatments.\\n10\\n, \\n11\\n If such actions are not taken, the burden of chronic diseases, which is referred to as an emerging prevalence of chronic diseases globally imposing an enormous economic and health burden will likely continue to rise,\\n10\\n, \\n11\\n which is alarming especially among the vulnerable and marginalized populations of Bangladesh with limited affordability for health services.\\n11\\n In addition, the emerging prevalence of chronic disease may also lead to a health system burden in terms of increasing healthcare utilization, treatment costs and chronic disease management.\\nIt is well established that chronic diseases are increasingly associated with the over‐utilization of healthcare services and a higher financial burden.\\n13\\n For example, a previous study documented that a higher number of chronic diseases was linked to an increased number of outpatient visits.\\n14\\n Therefore, adequate preventative services must be in place to reduce the social‐economic burden of chronic disease, thereby ensuring optimal use of health resources. However, the major challenges to ensure effective prevention and management of chronic diseases include social status, power gradients, racial/ethnic differences, poor accessibility and affordability of healthcare services.\\n2\\n, \\n6\\n, \\n15\\n, \\n16\\n, \\n17\\n Previous research has identified several factors associated with healthcare utilization, such as demographic and socioeconomic characteristics, type of healthcare providers and presence of chronic illnesses.\\n18\\n In addition, patients' preference for healthcare services depends on several factors, including personal preference, disease severity, economic capacity, the reputation of healthcare providers\\n18\\n as well as affordable costs associated with treatment.\\n19\\n Furthermore, short travel time to healthcare facilities, effective interactions with healthcare providers,\\n19\\n, \\n20\\n respectful service provider attitudes\\n21\\n and short waiting time\\n22\\n were positively associated with the patient preference for healthcare services. A recent discrete choice experiment study found that the availability of medicine and transport to the health facility were significant attributes of patient preference for healthcare services.\\n23\\n Notably, optimum healthcare utilization among chronically ill patients has a significant role in preventing and managing chronic diseases.\\nIn Bangladesh, a family spends an average of 11% of their total household budget on health care and half of the population spends 7% of their monthly per capita consumption expenditure on illness.\\n24\\n Understanding patients' preferences could help medical professionals and healthcare providers restructure the healthcare delivery model and ensure the quality of services. In addition to clinical guidelines, patients' preferences may also provide guidelines for the selection of treatment options. Patient preferences also help to inform clinical decisions where science has not yet been able to provide effective solutions to healthcare problems.\\n25\\n Therefore, information on patients' preferences for healthcare utilization in terms of healthcare expenditure is critical; without this knowledge, an effective national healthcare policy cannot be formulated.\\n26\\n Moreover, there is a current lack of evidence on the patients' preferences for healthcare services in chronic disease management in Bangladesh. Since no such study exists, the present study aimed to investigate patients' preference for healthcare services among adult patients in Bangladesh. Understanding patients' preferences are important in designing interventions aimed at reducing the burden of chronic diseases, which undermine economic growth and development and are increasingly becoming the leading cause of death.\",\n \"[SUBTITLE] Study design and data sources [SUBSECTION] This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\\n27\\n\\n\\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\\nThis study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\\n27\\n\\n\\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\\n[SUBTITLE] HIES survey sampling and sample size calculation [SUBSECTION] The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\\n27\\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\\n27\\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\\n27\\n using the following formula\\n\\n(1)\\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\\nThe sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\\n27\\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\\n27\\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\\n27\\n using the following formula\\n\\n(1)\\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\",\n \"This study followed an observational, cross‐sectional design. The data was extracted from the most recent nationally representative Household Income and Expenditure Survey (HIES), conducted during 2016–2017 by the Bangladesh Bureau of Statistics (BBS) (Supporting Information: Document 1). The HIES is a cross‐sectional survey conducted every 5 years since 1973 in Bangladesh with the objective of providing national estimates on income, expenditure and consumption, poverty, standard of living, health status and education.\\n27\\n\\n\\nIn the HIES survey, a semistructured questionnaire (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire) was used to collect information from the survey participants (adults aged 18 years or above) under nine modules: (1) household information, (2) education, (3) health‐illnesses and injuries, (4) economic activities and wage employment, (5) nonagricultural enterprises, (6) housing, (7) agriculture, (8) other assets and income and (9) consumption (Supporting Information: Appendix Table A1: HIES 2016–2017 Questionnaire). However, our analysis is based on the chronic disease‐related questions included in Module‐3: Health (Illnesses and Injuries) of the HIES (Supporting Information: Appendix Table A2). Therefore, we only used the indicators pertaining to chronic disease and health service utilization along with the sociodemographic characterizes of the participants (Supporting Information: Appendix Table A2). All health‐related information was self‐reported in the HIES. Respondents were asked to prioritize the chronic diseases they were suffering from in order of their importance. We selected the primary disease (i.e., principal diagnosis) based on the patients' experience of diseases in order of their importance. For instance, if one patient was diagnosed with three chronic conditions, they reported three diseases in order of importance (i.e., first, second and third importance). Therefore, a patient's first importance of disease was considered as the primary disease in the present study.\",\n \"The sample frame used in the selection of Primary Sample Units (PSUs) for the HIES 2016–2017 was based on the Census of Population and Housing 2011.\\n27\\n In this survey, eight administrative divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet) were included and stratified by three basic localities, being rural, urban and metropolitan areas (city corporations). Thus, these should be 8 divisions × 3 localities totalling 24 strata (8 × 3 = 24). However, the sampling frame (Census of Population and Housing 2011) did not contain three administrative divisions (i.e., Rangpur, Barisal and Sylhet). Additionally, the BBS included only four main city corporations (Dhaka, Chittagong, Khulna and Rajshahi) in the city corporation locality. PSUs were randomly selected from 20 strata (eight rural divisions, eight urban divisions and four metropolitan areas) for national representation. As the PSUs of HIES 2016–2017 were allocated at the district level, the sample design includes a total of 132 substrata: 64 rural, 64 urban and 4 metropolitan areas.\\n27\\n Sample size was calculated using the prevalence rate of the main indicator (poverty rate) or the coefficient of variation of per capita consumption or household consumption, which are the core indicators of the HIES 2016–2017. Each one was treated as a target variable for determining the sample size (Supporting Information: Document 1). The required sample size was calculated for each district and explained elsewhere,\\n27\\n using the following formula\\n\\n(1)\\nn=Zα2×CVSRS(y¯)r(Y¯)2×DEFF,\\nwhere, n was the required sample for allocation to each district to achieve a certain level in the accuracy statistic (r(Y®) = 10% relative standard error desired for the mean total household expenditure estimated at the district level associated with the targeted variable (y®);CVSRS(y®) was the coefficient of variation of the targeted variable [i.e., total household expenditure estimated at the national level] estimated under the assumption of simple random sampling; DEFF was the design effect of the target variable [i.e., the average design effect of the target variable across all districts]; and Zα2 [=1.96] was the critical value of a standard normal distribution with an α% (5%) significance level). Substituting all values in Equation (1), the required sample was 715 households for each district, nonetheless, 720 households were allocated to each district for practical consideration and to facilitate fieldwork and survey implementation management. A stratified, two‐stage cluster sampling technique was used in this survey. In the first stage, a total of 36 PSUs were drawn from each district by applying the probability proportional to size systematic sampling technique, using the number of households in each PSU as the measure of size. The 36 PSUs were randomly selected from rural, urban and city corporation substratum. The total number of PSUs included in the analysis was 2304 (64 districts × 36 PSU per district). In the second stage, 20 households were selected per PSU. Using this sampling technique, 46,076 households (2304 PUSs × 20 households per PSU) were included in the study analysis of HIES 2016–2017 data. Among the selected households, a total of 186,076 individuals were included.\",\n \"Data collection was accomplished between early April 2016 and late March 2017 through face‐to‐face interviews with the participants. A total of 128 enumeration teams, each comprised of five members (one supervising officer, two interviewers and two female facilitators) who were thoroughly trained, completed the data collection. A multilayered quality control measure was employed to ensure the quality of the collected data. The supervising officer of each team was responsible to verify all the questionnaires completed by the field staff before they were sent to the headquarter. Moreover, Deputy Directors of the District Statistical Offices as well as senior officials from the headquarter, frequently visited the sampled areas to ensure data quality (Supporting Information: Document 1).\\nThe survey team formulated an operational definition of each variable used in the questionnaire. For example, while collecting chronic disease‐related information, the enumerators were instructed to ask the respondents, ‘Have you suffered from any chronic illness/disability in the last 12 months or more?’ (if yes); then, participants were asked a second question: ‘What chronic illness/disability are you suffering from?’ with response options: (1) chronic fever, (2) Injuries/disability, (3) chronic heart disease, (4) respiratory diseases/asthma/bronchitis, (5) diarrhoea/dysentery, (6) gastric/ulcer, (7) blood pressure, (8) arthritis/rheumatism, (9) skin problem, (10) diabetes, (11) cancer, (12) kidney diseases, (13) liver diseases, (14) mental health, (15) paralysis, (16) ear/ENT problem, (17) eye problem or (18) other (specify) (Supporting Information: Appendix Table A1: HIES 2016–2017 questionnaire). Participants responded based on their disease diagnosis, experiences, symptoms of illness and course of treatment. To get valid information, the enumerators also probed where necessary, asked the respondents to show any relevant documents such as prescriptions and test reports, or explained to the respondents about chronic diseases using various case scenarios as outlined in the survey guidelines and instructions.\\n[SUBTITLE] Study population [SUBSECTION] All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\\nAll selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\",\n \"All selected household members were included in the survey. The participants were selected based on the HIES 2016–2017 survey protocol and following the same inclusion criteria: (1) an individual who had suffered from any chronic disease for the last 12 months or more and (2) patients who received any treatment due to chronic disease in the last 30 days. Based on these inclusion criteria, a total of 10,385 patients were selected for the present analysis (Figure 1).\\nDistribution of study participants\",\n \"[SUBTITLE] Outcome measures [SUBSECTION] Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\\nPatients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\\n[SUBTITLE] Chronic diseases and comorbid conditions [SUBSECTION] Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\\nParticipants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\\n[SUBTITLE] Covariates [SUBSECTION] Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\\nBased on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\",\n \"Patients' preference for chronic disease healthcare utilization was the primary outcome measure for this study. Participants responded to questions that asked them about their preference for selecting a healthcare provider: ‘why did you choose this provider?’ Response options were recoded as the quality of healthcare services, availability of medical doctors (e.g., availability of male or female doctors in a health facility), affordable healthcare services (i.e., acceptable costs), short distance to health facility (i.e., nearby) and others (e.g., referred by other providers).\",\n \"Participants were asked about chronic illness, ‘have you suffered from any chronic illness/disability in the last 12 months or more?’ The prevalence of chronic diseases was assessed based on this question. Individuals who suffered from chronic disease for the last 12 months (or more) and received any treatment due to chronic diseases were the only population considered in this study. The study population was diagnosed with chronic illnesses and conditions such as chronic heart disease, respiratory diseases, chronic gastric or ulcers, high blood pressure, arthritis or rheumatism, diabetes, chronic fever and other diseases. There is no gold‐standard method for measuring comorbidity status among patients with chronic diseases.\\n28\\n A previous review study identified that 21 separate approaches were executed to measure comorbidity status.\\n28\\n The selection of the approach depends on the study research questions, study design, data availability and population studied. The most straightforward approach to measuring comorbidity status is to investigate the distribution of individual comorbid conditions and to treat them independently and/or to combine them by summing the total number of conditions.\\n29\\n, \\n30\\n A single condition count approach was performed to measure comorbidity status in this study. The count of chronic health condition(s) was measured for each patient based on the number of disease exposures and who had been prescribed medication for their illness. It was counted as multiple responses if the patients had multiple chronic conditions. In addition, the principal diagnosis was defined based on patients diagnosed based on their reported first importance. Chronic comorbid conditions (principal diagnosis plus one, two or three or more comorbid conditions) were assessed in this study.\",\n \"Based on the ongoing literature and the authors' own expertise, we selected variables to address the study objective, including healthcare service‐related factors (e.g., type of healthcare service‐inpatient or outpatient, type of health facilities, waiting time for receiving healthcare services, out‐of‐pocket payment and location of consulted healthcare provider) and patients' sociodemographics factors (e.g., gender, age, marital status, education and employment) (Supporting Information: Appendix Table A2).\",\n \"The adjusted multinomial logistic regression model was used to identify the potential factors that had an association with the patient's preference for healthcare service. The dependent variable (chronic illness patient's preference for healthcare services) was characterized as a categorical measure in the regression model. An unadjusted analysis was performed using each of the explanatory variables for the following reasons: (1) primary screening of the selection of qualified variables, which were added in the adjusted model, (2) the χ\\n2 tests (or one‐way analysis where appropriate) were used to find the association between outcome and explanatory variables. However, the majority of the explanatory variables were categorical with two or more labels; therefore, an unadjusted analysis was performed to find the association between the outcome variable and different categories of explanatory variables. The explanatory variables were included in the adjusted model only if any label of the predictor was significant at a 5% or less risk level in the unadjusted model, which in turn was used to adjust for the associations of other potential confounders. For the explanatory variables, the category found to be least at risk of having patients' preferences for healthcare services related to chronic illness in the analysis was considered the reference category for constructing the relative risk ratio (RRR). Statistical significance was considered at a 5% risk level. All data analyses were undertaken using the statistical software Stata/SE 14 (StataCorp).\",\n \"Table 1 shows the participant's characteristics. Of the 10,385 patients with chronic diseases, approximately 50% were female, 51% were married and 70% were unemployed. The majority of the patients were aged between 18 and 45 (81%) years, and around one‐third (37%) had no formal education.\\nDistribution of participant's characteristics\\n\\nNote: p value was derived using χ\\n2 test or one‐way analysis of variance where appropriate.\\nAbbreviations: BDT, Bangladeshi Taka; CI, confidence interval; SD, standard deviation.\\n[SUBTITLE] The distribution of chronic disease and utilization of healthcare services [SUBSECTION] The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\\nThe most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\\n[SUBTITLE] Preferences for healthcare services for chronic illnesses [SUBSECTION] Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\\nApproximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\\n[SUBTITLE] Correlates of patient preference for chronic disease‐related healthcare services [SUBSECTION] Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\\nInfluencing factors on patient's preference of healthcare services\\n\\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\\nTable 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\\nInfluencing factors on patient's preference of healthcare services\\n\\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\",\n \"The most prevalent chronic diseases reported by the patients included gastric/ulcer (16%) followed by arthritis/rheumatism (14%) (Table 1). Most of the patients (65%) reported one diagnosed chronic condition, while approximately 24% of the patients had two, and 11% had three or more chronic conditions. Most of the patients (91%) reported utilizing outpatient healthcare services in the past 30 days. A high proportion of the patients visited general practitioner clinics (~40%), followed by pharmacy/dispensary (24%) and public hospitals (20%) for healthcare services. Furthermore, approximately 53% of the patients received services at rural health facilities. The average out‐of‐pocket healthcare expenditure for chronic illness in the last 30 days was 3848 BDT (~47.40 USD; 2017 price year).\",\n \"Approximately 30% of patients reported that they preferred quality healthcare services, whereas 28% preferred a short distance to the health facility. Furthermore, 22% of the patients preferred affordable healthcare costs as the main driver, and 11% expressed a preference for the doctor's availability. However, the scenario varied among different chronic illnesses and healthcare services (Table 1). For example, among the patients who received healthcare services from a pharmacy or dispensary, 49% preferred short distances to health facilities, and 32% reported their preference for healthcare affordability.\",\n \"Table 2 highlights the factors that influence a patient's choice of healthcare services. Among patients with heart disease, quality of healthcare services was 29% more preferred than affordability (RRR = 0.71; 95% confidence interval [CI]: 0.56–0.90; p < .001).\\nInfluencing factors on patient's preference of healthcare services\\n\\nNote: ***, ** and * denoted significance level at 0.1%, 1% and 5%, respectively.\\nAbbreviations: CI, confidence interval; ref, reference group; RRR, relative risk ratio.\\nSimilarly, diabetic patients exhibited a 38% higher preference for quality rather than affordability of healthcare services (RRR = 0.62; 0.47–0.81; p < .001). However, patients with three or more chronic comorbid conditions were 1.39 times more likely to prefer affordability over the quality of healthcare services (RRR = 1.39; 1.15–1.68; p < .001). Patients who received healthcare services from a public facility reported a higher preference for the availability of medical doctors or consultants (RRR = 2.93; 1.70–5.04; p < .001) than those receiving health care from private service providers. In addition, patients who received healthcare services from pharmacies or dispensaries were significantly more likely to prefer a short distance to a healthcare facility (RRR = 6.99; 4.80–9.86; p < .001) or affordability of healthcare services (RRR = 3.13; 2.25–4.36; p < .001), rather than the quality of the healthcare services and availability of doctors.\",\n \"Chronic diseases among adults are becoming a significant health concern in many LMICs, including Bangladesh. This is the first study to focus on patient preferences for healthcare services for chronic disease management, using a recent nationally representative HIES. This study provided evidence of patient preference for the utilization of healthcare services in chronic disease management. The top four dimensions of patient preference for health care in order of preference were quality of treatment (30.3%), short distance to health facilities (27.6%), affordability of health care (21.7%) and availability of doctors (11.0%). The severity of the disease and the characteristics of healthcare providers were the most important contributing factors in patients' preferences and decisions to seek healthcare services.\\nThis study showed that patients with heart disease were more likely to prefer quality health care than healthcare affordability or a short distance to a health facility. This is in agreement with previous studies, which have also documented that chronic heart disease patients are more likely to prefer quality healthcare services.\\n31\\n, \\n32\\n, \\n33\\n Several attributes might influence a patient's preference for quality healthcare services. For instance, heart disease may be associated with both acute episodes and high levels of long‐term adverse events (e.g., mortality and disability).\\n34\\n, \\n35\\n, \\n36\\n Such patients may require curative care as a component of a continuous, coordinated care model, covering both primary and postacute hospital care.\\n34\\n Moreover, due to the high disease burden and prolonged treatment, continuous use of healthcare resources (e.g., specialist consultation, diagnostic and medicine) may lead to a higher economic burden for households, individuals and society.\\n35\\n, \\n36\\n, \\n37\\n From a healthcare service providers' perspective, examining patients' experience and satisfaction with healthcare services could identify whether services are of an acceptable standard or highlight areas for potential quality improvement.\\n38\\n Existing research has revealed that a positive experience with healthcare professionals or other medical staff is directly related to patient preference for healthcare quality during hospitalization or course of treatment.\\n39\\n Other studies have indicated the importance of having confidence in the expertise and attentiveness of doctors or nurses,\\n40\\n, \\n41\\n healthcare provider's interpersonal communication skills and behaviours,\\n42\\n although shorter waiting times have also been mentioned/raised.\\n43\\n\\n\\nIn the present study, one of the most influential aspects of a patient's healthcare‐seeking behaviour was the healthcare provider's location; the shorter the distance to the healthcare facility the more likely the patient uptake of the healthcare services. It is plausible that patients may be unwilling or unable to travel long distances to access medical expertise or treatment, particularly if the nature of the chronic illness requires frequent appointments. This finding is supported by previous studies which reported that distance to healthcare facilities was a potential barrier to accessing healthcare services in LMICs.\\n44\\n, \\n45\\n Patients' educational level, employment status and chronic comorbid conditions were also identified as significant determinants influencing their healthcare‐seeking preferences. This may be due to patients with chronic illnesses being more likely to prioritize treatment, management and quality of life over other life choices they may face.\\n46\\n In addition, patients with multiple chronic conditions may require medical care from several healthcare providers across various locations throughout the year. This may be attributed to seeking quality care from experienced or specialist medical professionals, and superior diagnosis and treatment facilities, which has also been identified elsewhere.\\n47\\n\\n\\nAlthough this is the first study to investigate patient preference for healthcare services to manage chronic diseases among adults in Bangladesh, the extension and transferability of our findings to contexts beyond the study population should be handled with caution because of our study limitations. For instance, due to the cross‐sectional nature of the study, causality cannot be inferred. Furthermore, these findings may be subjected to some level of bias as data on the main variables of interest were self‐reported (i.e., illness, utilization and expenditure), thereby risking recall bias. However, the relatively short recall period (i.e., the last 30 days) of the household income and expenditure survey strives to reduce this potential bias.\\n[SUBTITLE] Implications for policy and practice [SUBSECTION] Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\\n48\\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\\n9\\n, \\n11\\n, \\n49\\n\\n\\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\\n11\\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\\n48\\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\\n11\\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.\\nOur study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\\n48\\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\\n9\\n, \\n11\\n, \\n49\\n\\n\\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\\n11\\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\\n48\\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\\n11\\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.\",\n \"Our study highlights the significance of patient preferences for healthcare utilization. This study provides timely findings to address health inequities linked to sociocultural and economic factors in Bangladesh. Understanding patients' preferences in chronic disease management is critical to achieving the Sustainable Development Goal (SDG) target 3.4, which focuses solely on reducing premature mortality from noncommunicable diseases by a third by 2030 relative when compared to 2015 as a baseline. Bangladesh is a signatory to the SDGs and the Colombo declaration (strengthening health systems to accelerate delivery of NCD‐related services at the primary healthcare level),\\n48\\n providing evidence that informs culturally competent NCD prevention and treatment approaches through tailored and responsive health financing, and expenditure policies will contribute significantly to achieving the SDG target 3.4. However, progress in implementing strategies to meet international targets related to NCDs remains slow.\\n9\\n, \\n11\\n, \\n49\\n\\n\\nAdditionally, Bangladesh lacks a national surveillance programme focused on chronic diseases and does not have any integrated community public health programme that regularly monitors chronic diseases.\\n11\\n An established disease management plan should consider the adequacy, accessibility, affordability and quality of services.\\n48\\n Most importantly, the factors influencing patient preferences for healthcare services may not have been considered when developing the chronic disease management policy.\\n11\\n Therefore, policymakers and public health practitioners should consider patient preferences regarding healthcare utilization in managing chronic diseases.\",\n \"Our study findings highlighted that patient preferences for healthcare services in chronic disease management were significantly associated with disease severity and healthcare providers' attributes. Therefore, policymakers and public health practitioners should consider patient preferences for managing chronic conditions within the national strategic frameworks to improve service provision, equitable distribution and uptake of the services.\",\n \"Rashidul A. Mahumud conceptualized the study, directed the data analysis and wrote most of the manuscript. Rashidul A. Mahumud performed the statistical analysis under the supervision of Andre M. N. Renzaho, Khorshed Alam and Asaduzzaman Khan. Marufa Sultana, Satyajit Kundu, Md. A. Rahman and Mostafa Kamal wrote portions of the manuscript. Rashidul A. Mahumud, Khorshed Alam, Andre M. N. Renzaho, Sabuj K. Mistry, Joseph K. Kamara, Md. G. Hossain, Mohammad A. Ali and Asaduzzaman Khan contributed to the study design, data interpretation and edited the manuscript. All named authors contributed to critically reviewing the revised initial draft of the manuscript. All authors read and approved the final draft.\",\n \"The authors declare no conflict of interest.\",\n \"This study has been prepared using a secondary de‐identified and fully anonymized cross‐sectional data set from the HIES conducted during 2016–2017 by the Bangladesh Bureau of Statistics. Therefore, no ethical approval was not required for this study.\",\n \"Supporting information.\\nClick here for additional data file.\\nSupporting information.\\nClick here for additional data file.\\nSupporting information.\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["background","methods",null,null,null,null,null,null,null,null,null,"results",null,null,null,"discussion",null,"conclusions",null,"COI-statement",null,"supplementary-material"],"string":"[\n \"background\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\",\n null,\n \"conclusions\",\n null,\n \"COI-statement\",\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["Bangladesh","chronic diseases","healthcare services","patient preferences"],"string":"[\n \"Bangladesh\",\n \"chronic diseases\",\n \"healthcare services\",\n \"patient preferences\"\n]"}}},{"rowIdx":362,"cells":{"title":{"kind":"string","value":"Loss of tetraspanin-7 expression reduces pancreatic β-cell exocytosis Ca"},"pmid":{"kind":"string","value":"36264270"},"background_abstract":{"kind":"string","value":"Tetraspanin-7 (Tspan7) is an islet autoantigen involved in autoimmune type 1 diabetes and known to regulate β-cell L-type Ca2+ channel activity. However, the role of Tspan7 in pancreatic β-cell function is not yet fully understood."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Histological analyses were conducted using immunostaining. Whole-body metabolism was tested using glucose tolerance test. Islet hormone secretion was quantified using static batch incubation or dynamic perifusion. β-cell transmembrane currents, electrical activity and exocytosis were measured using whole-cell patch-clamping and capacitance measurements. Gene expression was studied using mRNA-sequencing and quantitative PCR."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Tspan7 is expressed in insulin-containing granules of pancreatic β-cells and glucagon-producing α-cells. Tspan7 knockout mice (Tspan7y/- mouse) exhibit reduced body weight and ad libitum plasma glucose but normal glucose tolerance. Tspan7y/- islets have normal insulin content and glucose- or tolbutamide-stimulated insulin secretion. Depolarisation-triggered Ca2+ current was enhanced in Tspan7y/- β-cells, but β-cell electrical activity and depolarisation-evoked exocytosis were unchanged suggesting that exocytosis was less sensitive to Ca2+ . TSPAN7 knockdown (KD) in human pseudo-islets led to a significant reduction in insulin secretion stimulated by 20 mM K+ . Transcriptomic analyses show that TSPAN7 KD in human pseudo-islets correlated with changes in genes involved in hormone secretion, apoptosis and ER stress. Consistent with rodent β-cells, exocytotic Ca2+ sensitivity was reduced in a human β-cell line (EndoC-βH1) following Tspan7 KD."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Tspan7 is involved in the regulation of Ca2+ -dependent exocytosis in β-cells. Its function is more significant in human β-cells than their rodent counterparts."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Animals","Humans","Mice","Exocytosis","Glucose","Insulin","Insulin-Secreting Cells","Islets of Langerhans","Tetraspanins"],"string":"[\n \"Animals\",\n \"Humans\",\n \"Mice\",\n \"Exocytosis\",\n \"Glucose\",\n \"Insulin\",\n \"Insulin-Secreting Cells\",\n \"Islets of Langerhans\",\n \"Tetraspanins\"\n]"},"pmcid":{"kind":"string","value":"9828109"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Animals and isolation of pancreatic islets [SUBSECTION] All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\nAll animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\n[SUBTITLE] \nshRNA\n [SUBSECTION] Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\nAdenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\n[SUBTITLE] Human pseudo‐islets and EndoC‐βH1 cell line [SUBSECTION] Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n\nDonor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n\n[SUBTITLE] Insulin secretion [SUBSECTION] Insulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\nInsulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\n[SUBTITLE] Electrophysiology [SUBSECTION] Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n\nElectrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n\n[SUBTITLE] Histology [SUBSECTION] Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\nSections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\n[SUBTITLE] Glucose tolerance test [SUBSECTION] Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\nIntraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\n[SUBTITLE] quantitative PCR\n [SUBSECTION] RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses\nRNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses\n[SUBTITLE] \nRNA sequencing [SUBSECTION] Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\nHuman pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\n[SUBTITLE] Data analyses and statistics [SUBSECTION] Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\nData are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] \nTspan7 is expressed in pancreatic islets [SUBSECTION] Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\nTspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\n[SUBTITLE] Ablation of Tspan7 has limited impact on glucose tolerance [SUBSECTION] To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\nTo investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\n[SUBTITLE] Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells [SUBSECTION] In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\nIn β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\n[SUBTITLE] Reducing \nTSPAN7\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors [SUBSECTION] To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\nTo assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["What is already known?","What did this study find?","What are the implications of the study?","INTRODUCTION","Animals and isolation of pancreatic islets","\nshRNA\n","Human pseudo‐islets and EndoC‐βH1 cell line","Insulin secretion","Electrophysiology","Histology","Glucose tolerance test","quantitative PCR\n","\nRNA sequencing","Data analyses and statistics","\nTspan7 is expressed in pancreatic islets","Ablation of Tspan7 has limited impact on glucose tolerance","Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells","Reducing \nTSPAN7\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors"],"string":"[\n \"What is already known?\",\n \"What did this study find?\",\n \"What are the implications of the study?\",\n \"INTRODUCTION\",\n \"Animals and isolation of pancreatic islets\",\n \"\\nshRNA\\n\",\n \"Human pseudo‐islets and EndoC‐βH1 cell line\",\n \"Insulin secretion\",\n \"Electrophysiology\",\n \"Histology\",\n \"Glucose tolerance test\",\n \"quantitative PCR\\n\",\n \"\\nRNA sequencing\",\n \"Data analyses and statistics\",\n \"\\nTspan7 is expressed in pancreatic islets\",\n \"Ablation of Tspan7 has limited impact on glucose tolerance\",\n \"Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells\",\n \"Reducing \\nTSPAN7\\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors\"\n]"},"other_sections_texts":{"kind":"list like","value":["Tetraspanin‐7, a transmembrane protein widely expressed in pancreatic islet cells, was previously identified as an autoantigen in type 1 diabetes. Tetraspanin‐7 has been shown to regulate β‐cell L‐type Ca2+ channel activity.","Genetic ablation of tetraspanin‐7 in mice led to a significant reduction in the Ca2+ dependence of β‐cell exocytosis. This phenotype was also observed in human islet cells following ex vivo tetraspanin‐7 knockdown, alongside down‐regulation of the β‐cell exocytosis Ca2+ sensor, synaptotagmin‐7.","Tetraspanin‐7 expression is required for normal exocytotic machinery and appropriate exocytosis Ca2+ sensitivity in pancreatic β‐cells.","Tetraspanin‐7 (Tspan7) was recently established as an autoantigen in type 1 diabetes mellitus (T1D),\n1\n an autoimmune disease in which loss of pancreatic β‐cells leads to chronic hyperglycaemia.\n2\n Proteins of the tetraspanin superfamily, of which Tspan7 is a member, are characterised by the presence of four transmembrane domains: one short and one long extracellular loop, a short intracellular loop, and N‐terminal and C‐terminal cytoplasmic tails.\n3\n, \n4\n Tetraspanin proteins are typically organised in tetraspanin‐enriched microdomains on biological membranes\n5\n where they act as molecular facilitators, engaging both transmembrane and intracellular proteins to regulate diverse processes such as cell proliferation, differentiation, activation, adhesion, motility and signalling.\n5\n, \n6\n, \n7\n\n\nTspan7 is highly expressed in neuroendocrine tissue,\n8\n and loss of Tspan7 can lead to defects in neuronal morphogenesis and synaptic transmission,\n9\n similar to those described in people expressing a mutated form.\n10\n Additional functions in cell morphology have been reported in dendritic cells\n11\n and osteoclasts,\n12\n attributed to its role in actin cytoskeleton rearrangement. Within the pancreas, Tspan7 is localised to the pancreatic islets and was recently shown to function as an auxiliary protein of L‐type voltage‐gated Ca2+ channels (VGCCs), modulating β‐cell electrical excitability and exocytosis. Knocking‐down of Tspan7 in β‐cells led to higher Ca2+ influx (through L‐type VGCCs), cellular excitability and glucose‐stimulated insulin secretion.\n13\n\n\nIn this study, we characterised Tspan7y/− mouse metabolic phenotype, β‐cell excitability, glucose‐stimulated insulin secretion and cell exocytosis. These data provide further insights of the role of Tspan7 in regulating β‐cell intracellular Ca2+ dynamics and insulin secretion, and thus to determine the importance of Tspan7 in glucose tolerance.","All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).","Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.","Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n","Insulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).","Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n","Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).","Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.","RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses","Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.","Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.","Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.","To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.","In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).","To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test."],"string":"[\n \"Tetraspanin‐7, a transmembrane protein widely expressed in pancreatic islet cells, was previously identified as an autoantigen in type 1 diabetes. Tetraspanin‐7 has been shown to regulate β‐cell L‐type Ca2+ channel activity.\",\n \"Genetic ablation of tetraspanin‐7 in mice led to a significant reduction in the Ca2+ dependence of β‐cell exocytosis. This phenotype was also observed in human islet cells following ex vivo tetraspanin‐7 knockdown, alongside down‐regulation of the β‐cell exocytosis Ca2+ sensor, synaptotagmin‐7.\",\n \"Tetraspanin‐7 expression is required for normal exocytotic machinery and appropriate exocytosis Ca2+ sensitivity in pancreatic β‐cells.\",\n \"Tetraspanin‐7 (Tspan7) was recently established as an autoantigen in type 1 diabetes mellitus (T1D),\\n1\\n an autoimmune disease in which loss of pancreatic β‐cells leads to chronic hyperglycaemia.\\n2\\n Proteins of the tetraspanin superfamily, of which Tspan7 is a member, are characterised by the presence of four transmembrane domains: one short and one long extracellular loop, a short intracellular loop, and N‐terminal and C‐terminal cytoplasmic tails.\\n3\\n, \\n4\\n Tetraspanin proteins are typically organised in tetraspanin‐enriched microdomains on biological membranes\\n5\\n where they act as molecular facilitators, engaging both transmembrane and intracellular proteins to regulate diverse processes such as cell proliferation, differentiation, activation, adhesion, motility and signalling.\\n5\\n, \\n6\\n, \\n7\\n\\n\\nTspan7 is highly expressed in neuroendocrine tissue,\\n8\\n and loss of Tspan7 can lead to defects in neuronal morphogenesis and synaptic transmission,\\n9\\n similar to those described in people expressing a mutated form.\\n10\\n Additional functions in cell morphology have been reported in dendritic cells\\n11\\n and osteoclasts,\\n12\\n attributed to its role in actin cytoskeleton rearrangement. Within the pancreas, Tspan7 is localised to the pancreatic islets and was recently shown to function as an auxiliary protein of L‐type voltage‐gated Ca2+ channels (VGCCs), modulating β‐cell electrical excitability and exocytosis. Knocking‐down of Tspan7 in β‐cells led to higher Ca2+ influx (through L‐type VGCCs), cellular excitability and glucose‐stimulated insulin secretion.\\n13\\n\\n\\nIn this study, we characterised Tspan7y/− mouse metabolic phenotype, β‐cell excitability, glucose‐stimulated insulin secretion and cell exocytosis. These data provide further insights of the role of Tspan7 in regulating β‐cell intracellular Ca2+ dynamics and insulin secretion, and thus to determine the importance of Tspan7 in glucose tolerance.\",\n \"All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\\n14\\n, \\n15\\n\\nTspan7 knockout mice (Tspan7y/−)\\n9\\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\",\n \"Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\",\n \"Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\\n16\\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\\n17\\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\\n18\\n\\n\",\n \"Insulin secretion was measured in static incubations as previously described.\\n15\\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\",\n \"Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\\n19\\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\\n14\\n β‐cells were identified by electrophysiological fingerprinting.\\n20\\n\\n\",\n \"Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\",\n \"Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\\n15\\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\",\n \"RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\\nPrimers and types of assays used for quantitative PCR analyses\",\n \"Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\\n21\\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\",\n \"Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\",\n \"Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\\n13\\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\\n22\\n\\n\\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\",\n \"To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\\n9\\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\\n23\\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\",\n \"In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\\n13\\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\\n13\\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\\n24\\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\",\n \"To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\\n\\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\\n18\\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["What is already known?","What did this study find?","What are the implications of the study?","INTRODUCTION","METHODS","Animals and isolation of pancreatic islets","\nshRNA\n","Human pseudo‐islets and EndoC‐βH1 cell line","Insulin secretion","Electrophysiology","Histology","Glucose tolerance test","quantitative PCR\n","\nRNA sequencing","Data analyses and statistics","RESULTS","\nTspan7 is expressed in pancreatic islets","Ablation of Tspan7 has limited impact on glucose tolerance","Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells","Reducing \nTSPAN7\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors","DISCUSSION","CONFLICT OF INTEREST","Supporting information"],"string":"[\n \"What is already known?\",\n \"What did this study find?\",\n \"What are the implications of the study?\",\n \"INTRODUCTION\",\n \"METHODS\",\n \"Animals and isolation of pancreatic islets\",\n \"\\nshRNA\\n\",\n \"Human pseudo‐islets and EndoC‐βH1 cell line\",\n \"Insulin secretion\",\n \"Electrophysiology\",\n \"Histology\",\n \"Glucose tolerance test\",\n \"quantitative PCR\\n\",\n \"\\nRNA sequencing\",\n \"Data analyses and statistics\",\n \"RESULTS\",\n \"\\nTspan7 is expressed in pancreatic islets\",\n \"Ablation of Tspan7 has limited impact on glucose tolerance\",\n \"Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells\",\n \"Reducing \\nTSPAN7\\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors\",\n \"DISCUSSION\",\n \"CONFLICT OF INTEREST\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Tetraspanin‐7, a transmembrane protein widely expressed in pancreatic islet cells, was previously identified as an autoantigen in type 1 diabetes. Tetraspanin‐7 has been shown to regulate β‐cell L‐type Ca2+ channel activity.","Genetic ablation of tetraspanin‐7 in mice led to a significant reduction in the Ca2+ dependence of β‐cell exocytosis. This phenotype was also observed in human islet cells following ex vivo tetraspanin‐7 knockdown, alongside down‐regulation of the β‐cell exocytosis Ca2+ sensor, synaptotagmin‐7.","Tetraspanin‐7 expression is required for normal exocytotic machinery and appropriate exocytosis Ca2+ sensitivity in pancreatic β‐cells.","Tetraspanin‐7 (Tspan7) was recently established as an autoantigen in type 1 diabetes mellitus (T1D),\n1\n an autoimmune disease in which loss of pancreatic β‐cells leads to chronic hyperglycaemia.\n2\n Proteins of the tetraspanin superfamily, of which Tspan7 is a member, are characterised by the presence of four transmembrane domains: one short and one long extracellular loop, a short intracellular loop, and N‐terminal and C‐terminal cytoplasmic tails.\n3\n, \n4\n Tetraspanin proteins are typically organised in tetraspanin‐enriched microdomains on biological membranes\n5\n where they act as molecular facilitators, engaging both transmembrane and intracellular proteins to regulate diverse processes such as cell proliferation, differentiation, activation, adhesion, motility and signalling.\n5\n, \n6\n, \n7\n\n\nTspan7 is highly expressed in neuroendocrine tissue,\n8\n and loss of Tspan7 can lead to defects in neuronal morphogenesis and synaptic transmission,\n9\n similar to those described in people expressing a mutated form.\n10\n Additional functions in cell morphology have been reported in dendritic cells\n11\n and osteoclasts,\n12\n attributed to its role in actin cytoskeleton rearrangement. Within the pancreas, Tspan7 is localised to the pancreatic islets and was recently shown to function as an auxiliary protein of L‐type voltage‐gated Ca2+ channels (VGCCs), modulating β‐cell electrical excitability and exocytosis. Knocking‐down of Tspan7 in β‐cells led to higher Ca2+ influx (through L‐type VGCCs), cellular excitability and glucose‐stimulated insulin secretion.\n13\n\n\nIn this study, we characterised Tspan7y/− mouse metabolic phenotype, β‐cell excitability, glucose‐stimulated insulin secretion and cell exocytosis. These data provide further insights of the role of Tspan7 in regulating β‐cell intracellular Ca2+ dynamics and insulin secretion, and thus to determine the importance of Tspan7 in glucose tolerance.","[SUBTITLE] Animals and isolation of pancreatic islets [SUBSECTION] All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\nAll animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\n[SUBTITLE] \nshRNA\n [SUBSECTION] Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\nAdenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\n[SUBTITLE] Human pseudo‐islets and EndoC‐βH1 cell line [SUBSECTION] Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n\nDonor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n\n[SUBTITLE] Insulin secretion [SUBSECTION] Insulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\nInsulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\n[SUBTITLE] Electrophysiology [SUBSECTION] Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n\nElectrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n\n[SUBTITLE] Histology [SUBSECTION] Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\nSections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\n[SUBTITLE] Glucose tolerance test [SUBSECTION] Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\nIntraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\n[SUBTITLE] quantitative PCR\n [SUBSECTION] RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses\nRNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses\n[SUBTITLE] \nRNA sequencing [SUBSECTION] Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\nHuman pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\n[SUBTITLE] Data analyses and statistics [SUBSECTION] Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\nData are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.","All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\n14\n, \n15\n\nTspan7 knockout mice (Tspan7y/−)\n9\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).","Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.","Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\n16\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\n17\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\n18\n\n","Insulin secretion was measured in static incubations as previously described.\n15\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).","Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\n19\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\n14\n β‐cells were identified by electrophysiological fingerprinting.\n20\n\n","Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).","Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\n15\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.","RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\nPrimers and types of assays used for quantitative PCR analyses","Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\n21\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.","Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.","[SUBTITLE] \nTspan7 is expressed in pancreatic islets [SUBSECTION] Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\nTspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\n[SUBTITLE] Ablation of Tspan7 has limited impact on glucose tolerance [SUBSECTION] To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\nTo investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\n[SUBTITLE] Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells [SUBSECTION] In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\nIn β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\n[SUBTITLE] Reducing \nTSPAN7\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors [SUBSECTION] To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\nTo assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.","Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\n13\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\n22\n\n\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.","To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\n9\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\n23\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.","In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\n13\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\n13\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\n24\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).","To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\n\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\n18\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.","Autoantigens implicated in type 1 diabetes have been shown to play a role in the regulation of insulin secretion.\n25\n Tspan7, a recently identified T1D autoantigen, can interact with both intracellular and membrane proteins, regulating diverse cellular processes, in a variety of cell types.\n5\n While Tspan7 expression in β‐cells is well established,\n8\n, \n13\n, \n22\n its role in insulin secretion is only partially understood. Here, we show that, in addition to its role in regulating VGCCs,\n13\n adequate expression of Tspan7 is required to maintain Ca2+ sensitivity of β‐cell exocytosis, which is associated with appropriate expression of synaptotagmin 7. This correlation is found in both mouse and human β‐cells but exerts a more significant role in human β‐cell function.\nThe role of Tspan7 in regulating cellular functions has been studied in many different cell types\n5\n, \n6\n, \n9\n, \n10\n, \n12\n, \n23\n and this has been recently extended into the study of pancreatic β‐cells,\n13\n where it was found to reduce L‐type VGCC activity. Therefore, reducing Tspan7 expression in β‐cells would be expected to lead to higher Ca2+ influx, improving glucose tolerance through enhanced GSIS (a Ca2+‐dependent process). Indeed, in Tspan7y/− β‐cells, depolarisation‐triggered Ca2+ current was augmented (Figure 4a; Figure S3a) and this may contribute to the trend of increased action potential peak, decreased inter‐spike plateau potential and lower firing frequency at high glucose (Figure 3b–d), similar to the effect of an L‐type VGCC agonist BAYK 8644.\n26\n However, this increment (~ + 3pA/pF at 0 mV) did not significantly change β‐cell action potential (Figure S1) and thus, may not be sufficient to impact on insulin secretion. This is confirmed by the normal GSIS observed in Tspan7y/− mice, both in vivo (Figure S1) and in vitro (Figure 4f). Similarly, GSIS was not significantly altered in human pseudo‐islets transduced with TSPAN7 shRNA. This differs from the previous report by Dickerson et al.\n13\n and may result from the different experimental paradigms used across the two studies (i.e., 2D single cell culture or 3D pseudo‐islets). Pseudo‐islets, formed by re‐aggregating dispersed islet cells, is an approximation of the primary intact islet, preserving a more physiological environment that includes intra‐islet paracrine signalling. Pseudo‐islet function is comparable with primary human islets\n17\n and allows efficient genetic manipulation using transduction, as highlighted in the present study (Figure 5a).\nThe stimulus‐secretion coupling of β‐cells involves exocytosis that depends on Ca2+ influx through the L‐type VGCC.\n14\n However, the observation that depolarisation‐triggered insulin secretion was unaffected in Tspan7y/− islets (Figure 4f; Figure S4) and slightly reduced in TSPAN7 KD human pseudo‐islets (Figure 5b) suggests reduced Tspan7/TSPAN7 may also affect exocytotic machinery. By correlating the transmembrane Ca2+ influx with high‐resolution exocytosis measurements, it is clear that exocytosis Ca2+ sensitivity is impaired when the level of tetraspanin‐7 in β‐cells is low. This effect could be in part attributed to reduced expression of Syt7, the major β‐cell exocytosis Ca2+ sensor.\n27\n Down‐regulation of the Syt7 following tetraspanin‐7 ablation/knockdown may be due to: (1) a coordinated expressions of these genes, or (2) Ca2+‐dependent changes in gene expression (secondary to augment VGCC activity, reviewed in\n28\n). Either possibility suggests that expression of tetraspanin‐7 may be required for appropriate assembly of exocytotic machinery in β‐cells. It is therefore likely that TSPAN7 plays a more significant role in insulin secretion in human β‐cells, in which exocytosis is less dependent on L‐type VGCC (more P/Q‐type VGCC dependent\n29\n). This notion is consistent with the observation that insulin secretion induced by 20 mM K+ and GSIS was reduced in TSPAN7 KD pseudo‐islets and EndoC‐βH1 cells respectively (Figure 5b,g). It is interesting to note that the impact of TSPAN7 KD on pseudo‐islets GSIS is less pronounced. This highlights the difference between the primary islets or reconstituted organoids and human cell lines. Whereas the primary human islets (and the pseudo‐islets) are comprised of different type of endocrine cells, EndoC‐βH1 cells are a single β‐cell population, lacking of the local paracrine cross‐talks that are important for appropriate GSIS. For example, it has been reported recently that α‐cells facilitate GSIS through glucagon/GLP‐1‐dependent signalling pathways.\n30\n Therefore, it is possible that defects in exocytosis (as the consequence of TSPAN7 KD) could exert a stronger effect in EndoC‐βH1 cells. It will be interesting to test whether a significantly reduced GSIS can be also observed in TSPAN7 KD pseudo‐islet composed by pure human primary β‐cells.\n\nTSPAN7 knockdown related transcription change is not restricted to exocytotic pathways. There is a small but significant up‐regulation of genes involved in apoptosis. It is possible this is due to augmented VGCC activity and secondary to elevated cytosolic Ca2+, which is pro‐apoptotic (reviewed in24). However, the degree of up‐regulation in apoptosis‐related genes may not be sufficient to reduce β‐cell mass in either acute or chronic TSPAN7 loss of function, as evidenced by the normal islet insulin content in shTSPAN7 transduced EndoC‐βH1 cells and in aged Tspan7y/− mice.\nIn summary, in addition to its inhibitory role on β‐cell VGCCs, we found that Tspan7 is required for appropriate expression of genes involved in cell survival and exocytosis. The non‐apparent metabolic phenotype and normal GSIS found in Tspan7y/− mice may be the result of a combined effect of enhanced VGCC activity and reduced sensitivity of exocytosis to Ca2+ in β‐cells.","The authors do not have any relevant conflict of interest to disclose.","\nFigure S1.\n\nClick here for additional data file.\n\nFigure S2.\n\nClick here for additional data file.\n\nFigure S3.\n\nClick here for additional data file.\n\nFigure S4.\n\nClick here for additional data file.\n\nFigure S5.\n\nClick here for additional data file.\n\nData S1.\n\nClick here for additional data file."],"string":"[\n \"Tetraspanin‐7, a transmembrane protein widely expressed in pancreatic islet cells, was previously identified as an autoantigen in type 1 diabetes. Tetraspanin‐7 has been shown to regulate β‐cell L‐type Ca2+ channel activity.\",\n \"Genetic ablation of tetraspanin‐7 in mice led to a significant reduction in the Ca2+ dependence of β‐cell exocytosis. This phenotype was also observed in human islet cells following ex vivo tetraspanin‐7 knockdown, alongside down‐regulation of the β‐cell exocytosis Ca2+ sensor, synaptotagmin‐7.\",\n \"Tetraspanin‐7 expression is required for normal exocytotic machinery and appropriate exocytosis Ca2+ sensitivity in pancreatic β‐cells.\",\n \"Tetraspanin‐7 (Tspan7) was recently established as an autoantigen in type 1 diabetes mellitus (T1D),\\n1\\n an autoimmune disease in which loss of pancreatic β‐cells leads to chronic hyperglycaemia.\\n2\\n Proteins of the tetraspanin superfamily, of which Tspan7 is a member, are characterised by the presence of four transmembrane domains: one short and one long extracellular loop, a short intracellular loop, and N‐terminal and C‐terminal cytoplasmic tails.\\n3\\n, \\n4\\n Tetraspanin proteins are typically organised in tetraspanin‐enriched microdomains on biological membranes\\n5\\n where they act as molecular facilitators, engaging both transmembrane and intracellular proteins to regulate diverse processes such as cell proliferation, differentiation, activation, adhesion, motility and signalling.\\n5\\n, \\n6\\n, \\n7\\n\\n\\nTspan7 is highly expressed in neuroendocrine tissue,\\n8\\n and loss of Tspan7 can lead to defects in neuronal morphogenesis and synaptic transmission,\\n9\\n similar to those described in people expressing a mutated form.\\n10\\n Additional functions in cell morphology have been reported in dendritic cells\\n11\\n and osteoclasts,\\n12\\n attributed to its role in actin cytoskeleton rearrangement. Within the pancreas, Tspan7 is localised to the pancreatic islets and was recently shown to function as an auxiliary protein of L‐type voltage‐gated Ca2+ channels (VGCCs), modulating β‐cell electrical excitability and exocytosis. Knocking‐down of Tspan7 in β‐cells led to higher Ca2+ influx (through L‐type VGCCs), cellular excitability and glucose‐stimulated insulin secretion.\\n13\\n\\n\\nIn this study, we characterised Tspan7y/− mouse metabolic phenotype, β‐cell excitability, glucose‐stimulated insulin secretion and cell exocytosis. These data provide further insights of the role of Tspan7 in regulating β‐cell intracellular Ca2+ dynamics and insulin secretion, and thus to determine the importance of Tspan7 in glucose tolerance.\",\n \"[SUBTITLE] Animals and isolation of pancreatic islets [SUBSECTION] All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\\n14\\n, \\n15\\n\\nTspan7 knockout mice (Tspan7y/−)\\n9\\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\\nAll animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\\n14\\n, \\n15\\n\\nTspan7 knockout mice (Tspan7y/−)\\n9\\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\\n[SUBTITLE] \\nshRNA\\n [SUBSECTION] Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\\nAdenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\\n[SUBTITLE] Human pseudo‐islets and EndoC‐βH1 cell line [SUBSECTION] Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\\n16\\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\\n17\\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\\n18\\n\\n\\nDonor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\\n16\\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\\n17\\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\\n18\\n\\n\\n[SUBTITLE] Insulin secretion [SUBSECTION] Insulin secretion was measured in static incubations as previously described.\\n15\\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\\nInsulin secretion was measured in static incubations as previously described.\\n15\\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\\n[SUBTITLE] Electrophysiology [SUBSECTION] Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\\n19\\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\\n14\\n β‐cells were identified by electrophysiological fingerprinting.\\n20\\n\\n\\nElectrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\\n19\\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\\n14\\n β‐cells were identified by electrophysiological fingerprinting.\\n20\\n\\n\\n[SUBTITLE] Histology [SUBSECTION] Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\\nSections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\\n[SUBTITLE] Glucose tolerance test [SUBSECTION] Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\\n15\\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\\nIntraperitoneal (i.p.) glucose tolerance test was performed as previously described.\\n15\\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\\n[SUBTITLE] quantitative PCR\\n [SUBSECTION] RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\\nPrimers and types of assays used for quantitative PCR analyses\\nRNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\\nPrimers and types of assays used for quantitative PCR analyses\\n[SUBTITLE] \\nRNA sequencing [SUBSECTION] Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\\n21\\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\\nHuman pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\\n21\\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\\n[SUBTITLE] Data analyses and statistics [SUBSECTION] Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\\nData are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\",\n \"All animal experiments were conducted in accordance with the UK Animals Scientific Procedures Act (1986) and the University of Oxford ethical guidelines. Mouse islets were isolated as previously described.\\n14\\n, \\n15\\n\\nTspan7 knockout mice (Tspan7y/−)\\n9\\n were gifts from Dr Luca Murru (CNR Institute of Neuroscience, Milano, Italy). Hormone secretion studies and the electrophysiology experiments were performed on islets isolated from Tspan7y/− mice and their wildtype littermates (control).\",\n \"Adenoviruses expressing shRNA for silencing of human TSPAN7 (Ad‐ShTSPAN7, SKU# shADV‐226,651) or scrambled shRNA control (Ad‐Scramble, SKU# 1122) together with eGFP under a CMV promoter were purchased from Vector Biolabs.\",\n \"Donor organs were obtained with written consent and research ethics approval at the University of Alberta, and human islets were isolated as previous described\\n16\\n at the University of Alberta. All human islet information is listed in Table S1. Islets were dispersed into a single‐cell suspension using trypsin (TryLE Express, Gibco) and transduced with Ad‐ShTSPAN7 or Ad‐Scramble before forming human pseudo‐islets using centrifugal‐forced aggregation as previously described.\\n17\\n EndoC‐βH1 was provided by Endocell and cultured as previously described.\\n18\\n\\n\",\n \"Insulin secretion was measured in static incubations as previously described.\\n15\\n Briefly, groups of 10–20 isolated islets (number depended on islet availability) were pre‐incubated for 30 min at 37°C in Krebs‐Ringer buffer (KRB) consisting of (mM): 140 NaCl, 3.6 KCl, 0.5 MgSO4, 2.6 CaCl2, 0.5 NaH2PO4, 2 NaHCO3, 1 glucose and 5 HEPES (pH = 7.4 with NaOH), before being incubated for 60 min at 37°C in 1 ml of KRB supplemented as indicated in the legends. Immediately after incubation, an aliquot of the medium was removed for determination of insulin concentration using an ELISA kit (Mercodia, Sweden).\\nFor dynamic insulin secretion measurements, groups of 50 pseudo‐islets were perifused with KRB containing 1 or 16.7 mM glucose with or without 20 mM K+ (as indicated) at the rate of 100 μl/min, using a perifusion system (Biorep, FL, USA). Perfusate was collected every 2 min and insulin concentration was determined using the Alpco stellux human ELISA kits (Salem, NH).\\nThe perifusion experiments were conducted at the University of Alberta and the mRNAs of pseudo‐islets were then extracted using TRIzol (ThermoFisher) before shipped to Oxford for transcriptomic study.\\nEndoC‐βH1 cells were sequentially exposed for 20 min to 1 and 20 mM glucose containing KRB. Supernatants were collected at the end of each incubation and insulin cellular contents were extracted using acid ethanol. Insulin concentrations were determined using an ELISA kit (Mercodia, Sweden).\",\n \"Electrophysiological measurements were performed using an EPC‐10 patch‐clamp amplifier and Pulse software (version 8.80, HEKA Electronics, Germany). Electrical activity and KATP‐channel conductance were measured using perforated patch‐clamping technique as previously described\\n19\\n; membrane currents and changes in cell capacitance were recorded using voltage clamp in combination with capacitance measurements, as described in.\\n14\\n β‐cells were identified by electrophysiological fingerprinting.\\n20\\n\\n\",\n \"Sections of formalin‐fixed paraffin‐embedded tissue were de‐waxed and subjected to epitope retrieval in a microwave pressure cooker in 10 mM citric acid pH 6.0, 0.05% Tween 20. After blocking with 2.5% normal horse serum, sections were incubated overnight (4°C) with rabbit anti‐Tspan7 antibody (Anti‐TM4SF2, 1:50, Sigma‐Aldrich). Antibody labelling was detected with the VectaFluor™ Excel Amplified Anti‐Rabbit IgG, DyLight™ 488 Antibody Kit (Vector Laboratories). Slides were counterstained with guinea pig anti‐insulin (1:500, Dako) and mouse anti‐glucagon (1:500, Sigma) antibodies followed by labelling with goat anti‐guinea pig‐Alexa 633 (1:100) and goat anti‐mouse Alexa 555 (1:100) secondary antibodies (ThermoFisher). Slides were visualised by confocal microscopy.\\nFor electron microscopy, isolated human islets were fixed in 2.5% glutaraldehyde in PBS, postfixed in 1% OsO4, dehydrated and embedded in Spurr's resin. Sections were cut onto nickel grids before being immunolabelled using anti‐TM4SF2 (1:20) and protein A gold particles (15 nm, Biocell), guinea pig anti‐insulin antibody (DAKO, 1:500) followed by anti‐guinea pig gold 10 nm (British Biocell International), and mouse anti‐glucagon (1:500, Sigma) followed by anti‐mouse pig gold 10 nm (British Biocell International). Sections were viewed with a Joel 1010 microscope (accelerating voltage 80 kV).\\nTo establish the specificity of Tspan7 antibody, immunostaining was conducted using a stable Tspan7 KD INS‐1 cell line (cell line generation is described in Data S1). Tspan7 KD INS‐1 or control cells were fixed in 4% PFA, permeabilised using 0.1% Triton X‐100 and blocked with 5% normal swine serum. Tspan7 was detected using rabbit anti‐Tspan7 primary antibody (anti‐TM4SF2, 1:50, Sigma‐Aldrich; overnight incubation, 4°C), and with goat anti‐rabbit IgG Alexa Fluor 568 secondary antibody (Life Technologies). Nucleus were stained with RedDot2 (BioTium; 1:200 dilution). Slides were visualised by confocal microscopy (BioRad).\",\n \"Intraperitoneal (i.p.) glucose tolerance test was performed as previously described.\\n15\\n Briefly, the animals were individually fasted for 6 h before i.p. injected with a bolus of glucose (2 g/kg body weight). Blood glucose were monitored for 90 min using a glucometer by tail vein sampling.\",\n \"RNA was extracted from human pseudo‐islets using TRIzol Reagent (Life Technologies, 15,596,026) according to the manufacturer's instructions. cDNA was generated using the GoScript Reverse Transcription Kit (Promega, A5000). qPCR was performed using 20 ng of cDNA, and TaqMan Gene Expression Master Mix (Life Technologies Ltd, 4,369,016) or SYBR™ Green PCR Master Mix (Applied Biosystems, 4,309,155) with the gene expression assays or primers as detailed in Table 1. Gene expression was determined using the ΔΔCT method by normalising to HouseKeeping Genes (HKG) and to the level of expression of each target detected in Ad‐Scramble transduced cells. Data are presented as percentage of control.\\nPrimers and types of assays used for quantitative PCR analyses\",\n \"Human pseudo‐islets were preserved in Trizol before their total RNA was extracted using Quick‐RNA MicroPrep kit (R1050 Zymo Research). RNA libraries for RNA‐seq were prepared using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina following manufacturer's protocols. Sequencing was performed on NextSeq 550. Single‐end reads were splice‐aligned to a human genome (GRCh38) using GSNAP. FeatureCounts was used to assign reads to exons thus eventually getting counts per gene. EdgeR package of R\\n21\\n was used to perform differential analysis between the conditions (shRNA vs control), while controlling for the confounding effect from the donor. Across‐samples normalisation was performed using the TMM normalisation method.\",\n \"Data are expressed as mean value ± SEM. p‐values less than 0.05 were considered significant. Statistical analyses were performed using OriginPro 2017 either by Student's t‐test, paired comparison or two‐way ANOVA and Tukey post hoc test. Ns represent the number of cells for electrophysiological analyses, and of independent mice or cell passages for hormone assays and expression data.\",\n \"[SUBTITLE] \\nTspan7 is expressed in pancreatic islets [SUBSECTION] Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\\n13\\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\\n22\\n\\n\\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\\nTspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\\n13\\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\\n22\\n\\n\\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\\n[SUBTITLE] Ablation of Tspan7 has limited impact on glucose tolerance [SUBSECTION] To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\\n9\\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\\n23\\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\\nTo investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\\n9\\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\\n23\\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\\n[SUBTITLE] Exocytosis sensitivity to Ca2+ was reduced in Tspan7y/− β‐cells [SUBSECTION] In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\\n13\\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\\n13\\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\\n24\\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\\nIn β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\\n13\\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\\n13\\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\\n24\\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\\n[SUBTITLE] Reducing \\nTSPAN7\\n expression in human islets changes gene expression of islet exocytosis Ca2+ sensors [SUBSECTION] To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\\n\\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\\n18\\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\\nTo assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\\n\\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\\n18\\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\",\n \"Tspan7 expression in pancreatic islets was evaluated using immunohistochemistry. In both human and mouse pancreatic islets, Tspan7 was found to colocalise with insulin and glucagon, confirming its expression in β‐ and α‐cells\\n13\\n (Figure 1a). We observed enriched immunogold labelling of Tspan7 in large dense‐core vesicle of human α‐ and β‐cells (Figure 1b,c), consistent with a previous report that Tspan7 was present in the cytoplasm of islet cells with a distribution pattern similar to islet hormones.\\n22\\n\\n\\nTetraspanin‐7 is expressed in pancreatic islets of Langerhans. (a) Immunofluorescent staining of human (upper) and mouse (lower) pancreas sections. Tetraspanin‐7 (TSPAN7 or Tspan7, red) is detected in β (Insulin, green)‐ and α (Glucagon, grey)‐cells. Scale bar: 30 μm. (b, c) Immunogold labelling of TSPAN7 (15 nm gold particles, black arrows), in glucagon (b) and insulin (c) containing vesicles in human islets. Insulin was labelled with 10 nm gold particles (white arrows). Scale bar: 500 nm.\",\n \"To investigate the role of Tspan7 in islet function and systemic metabolism, a Tspan7 knockout mouse model (Tspan7y/−)\\n9\\n was used. Reduced body weight and ad libitum plasma glucose was observed in Tspan7y/− mice (Figure 2a,b), but this was not linked to changes in plasma insulin (measured as insulin and C‐peptide, Figure 2c,d). These effects may instead be attributable to the neurological manifestations of Tspan7 ablation, including depressive behaviour.\\n23\\n We observed no change in circulating proinsulin levels (Figure 2e), suggesting normal insulin processing in Tspan7y/− β‐cells. Furthermore, when subjected to intra‐peritoneal glucose tolerance tests (IPGTTs), no difference in glucose tolerance (Figure 2f) or glucose‐induced insulin secretion (GSIS, Figure S1) was observed between Tspan7y/− and control mice.\\nIn vivo metabolic phenotyping of Tspan7y/− mouse model. (a–e) Ad libitum bodyweight (a), and plasma concentrations of glucose (b), insulin (c), C‐peptide (d) and proinsulin measured in Tspan7y/− mice (red) and age‐matched (11.3 ± 0.3 wks) litter mate control (black). *p < 0.05 versus control and **p < 0.01 versus control. (f) Plasma glucose concentrations of Tspan7y/− mice (red) and age‐matched litter mate control (black) measured during intraperitoneal glucose tolerance tests. Glucose bolus was injected at 0 min. N = 5 for control and 6 for Tspan7y/−. Data are presented as mean value ± SEM.\",\n \"In β‐cells, glucose metabolism increases the intracellular ATP/ADP ratio, resulting in closure of ATP‐sensitive K+ channels (KATP‐channels) inducing membrane depolarisation and action potential (AP) firing. APs open voltage‐gated Ca2+ channels (VGCCs), enabling influx of Ca2+ to trigger Ca2+‐dependent exocytosis, culminating in insulin secretion. To interrogate the impact of Tspan7 ablation on β‐cell function at a single‐cell level, we phenotyped Tspan7y/− β‐cell excitability, transmembrane Ca2+ currents and exocytosis using electrophysiological techniques. In the presence of 1 mM glucose, Tspan7y/− and control β‐cells were both repolarised (Tspan7y/− vs. control: −80 ± 1 mV vs. −80 ± 2 mV; p = 0.4) and electrically silent. Elevating extracellular glucose or blocking the KATP channels (using tolbutamide, 200 μM) induced membrane depolarisation and firing of APs in β‐cells of both genotypes (Figure 3a). In the presence of 10 mM glucose, although the Tspan7y/− β‐cells were less depolarised (Tspan7y/− vs. control: −57 ± 4 mV vs. −46 ± 5 mV; p < 0.05), no difference in AP peak or firing frequency was observed across the two genotypes (Figure 3b–d). There was no apparent difference in AP halfwidth or spike duration, either (Figure S2). Further increasing extracellular glucose to 20 mM or inhibiting KATP channels with tolbutamide exerted similar effects on the electrical activity of both Tspan7y/− and control β‐cells (Figure 3b–d). However, Tspan7y/− β cells responded to 20 mM glucose stimulation faster than the control (Tspan7y/− vs. control: 2.4 ± 0.3 min vs. 3.85 ± 0.3 min, p < 0.05). These marginal changes in Tspan7y/− β‐cell electrical activity are not due to any apparent changes in KATP‐channel conductivity or glucose sensitivity (Figure 3e,f).\\nElectrical activity and KATP‐channel conductance measured in Tspan7y/− β‐cells. (a) Membrane potential recordings of control (upper trace, black) and Tspan7y/− (lower trace, red) β‐cells within intact islets in response to changes in extracellular glucose (1 mM, 10 mM and 20 mM) or application of 200 μM tolbutamide, as indicated). (b) Summary of control (black) and Tspan7y/− (red) β‐cell membrane potential measured at conditions indicated. N = 7 for control and 14 for Tspan7y/− β‐cells at 1 mM glucose; N = 7 for control and 11 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 6 for Tspan7y/− β‐cells when 200 μM tolbutamide was applied. *p < 0.05 versus control. (c, d), as in (b) but summarise the peak voltage (c) and firing frequency (d) of action potentials measured in control (black) and Tspan7y/− (red) β‐cells at conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 3 for Tspan7y/− β‐cells when tolbutamide was applied. (e) Summary of KATP‐channel conductance measured in control (black) and Tspan7y/− (red) β‐cells under the conditions indicated. N = 4 for control and 7 for Tspan7y/− β‐cells at 1 mM and 10 mM glucose; N = 4 for control and 6 for Tspan7y/− β‐cells at 20 mM glucose and N = 3 for control and 5 for Tspan7y/− β‐cells when tolbutamide was applied. (f) Examples of KATP currents measured in control (black, middle trace) and Tspan7y/− (red, bottom trace) β‐cells. KATP currents were triggered by ±10 mV excursions from −70 mV (200 ms, top) in voltage‐clamped β‐cells under the conditions indicated. All data are presented as mean value ± SEM.\\nTransient knockdown of Tspan7 in β‐cells was recently shown to increase Ca2+ influx through L‐type VGCC.\\n13\\n We assessed whether a similar phenotype can be observed in β‐cells of mice with a constitutive Tspan7 knockout. As shown in Figure 4a,b, depolarisation triggered a significantly larger Ca2+ current in Tspan7y/− β‐cells than in the control (p < 0.01, ~160% at 0 and 10 mV), consistent with Dickerson et al.\\n13\\n As Ca2+ triggers exocytosis, we tested whether enhanced VGCC activity led to higher exocytosis in Tspan7y/− β‐cells using capacitance measurements. Exocytosis (measured as increases in cell capacitance) was trigged by a series of depolarising pulses from −70 mV to 0 mV with progressively longer durations (from 10 ms to 800 ms). Unexpectedly, exocytosis was comparable between Tspan7y/− and control β‐cells at all pulse durations (Figure 4c,d). We correlated the Ca2+ charges (total influx Ca2+ ions, QCa2+) with the corresponding exocytosis elicited by depolarisation (Figure 4e). Whereas exocytosis could be triggered by 0.15pC/pF QCa2+ in control β‐cells, it required > ~ 0.2pC/pF in Tspan7y/− β‐cells, suggesting a reduced exocytotic Ca2+sensitivity.\\nCa2+ current and exocytosis in Tspan7y/− β‐cells. Tspan7 loss of function significantly increases Ca2+ current density in β cells (a, b). (a) Representative traces of control (black) and Tspan7y/− (red) β‐cell Ca2+ current elicited by a 100‐ms depolarisation from −70 mV to 0 mV. (b) Summary of control (black) and Tspan7y/− (red) β‐cell Ca2+ current density in relationship to membrane voltages. N = 7 for control and 5 for Tspan7y/− β‐cells, **p < 0.01 versus control, paired comparison and Tukey. (c) Examples of exocytosis induced by membrane depolarisations from −70 mV to 0 mV (at the duration of 100 ms, 200 ms, 300 ms, 500 ms and 800 ms as indicated above the traces) of control (black) and Tspan7y/− (red) β‐cells. (d) Summary of control (black) and Tspan7y/− (red) β‐cell exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 11 cells for control and 14 cells for Tspan7y/− β‐cell. (e) Control (black) and Tspan7y/− (red) β‐cell charges‐exocytosis relationship for pulses up to 100 ms. Corresponding durations of the pulses are as labelled. N = 7 cells for control and 5 cells for Tspan7y/− β‐cells. (f) Insulin secretion measured from batch‐incubated control (black) and Tspan7y/− (red) islets in response to glucose or tolbutamide (concentrations as indicated). Islets were isolated from mice aged 20 weeks and N ≥ 3 for control and Tspan7y/− islets. **p < 0.01 and ***p < 0.001 versus 1 mM glucose alone of the same genotype. All data are presented as mean value ± SEM.\\nGiven the limited changes in electrical activity and exocytosis in Tspan7y/− β‐cells, we reasoned that knocking‐out Tspan7 would have little effect on glucose/tolbutamide‐stimulated insulin secretion. Indeed, no significant difference was observed in insulin secretion, stimulated by 10 mM glucose or 0.2 mM tolbutamide, from islets of control and Tspan7y/− mice (Figure 4f). We reasoned that enhanced VGCC activity augments the concentration of cytosolic Ca2+, and thus may induce apoptosis,\\n24\\n affecting islet β cell mass in aged animals. However, no significant difference in GSIS, insulin secretion induced by 70 mM K+ or islet insulin content was observed in control and Tspan7y/− islets isolated from 1‐year‐old mice (Figure S4).\",\n \"To assess the role of tetraspanin‐7 in human β‐cell function, we utilised an adenovirus encoding shRNATSPAN7 (Ad‐ShTSPAN7) to knockdown (KD) human TSPAN7 expression. Islet cells transduced with Ad‐ShTSPAN7, or the control virus (with scrambled RNA), were used to form pseudo‐islets. RNA sequencing showed a marked reduction in TSPAN7 expression (92.65%) in the TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −3.765, p\\nadj = 6.42 E‐13) (Figure 5a), confirming successful genetic manipulation. In total, 1030 genes (1000 are protein‐coding genes) were down‐regulated and 395 genes (372 are protein‐coding genes) were up‐regulated in TSPAN7‐KD pseudo‐islets. Pathway analyses showed up‐regulation in apoptosis gene ontology pathway (p\\nadj = 4.08 E−7); and down‐regulation of genes involved in the secretory pathway and transport regulation (p\\nadj = 5.62 E−9 and 4.22 E−08 respectively). Interestingly, β‐cell expression of the exocytosis Ca2+ sensor SYT7 was reduced in TSPAN7‐KD pseudo‐islets (log2[Fold‐change] = −0.39, p\\nadj = 3.89 E−3). GSIS of the pseudo‐islets were tested using dynamic perifusion experiments (Figure 5b). Whereas insulin secretion at basal or 16.7 mM glucose was comparable between the TSPAN7‐KD and control pseudo‐islets, 20 mM K+‐stimulated insulin secretion was significantly lower in TSPAN7‐KD pseudo‐islets.\\n\\nTSPAN7 KD affects exocytosis from human pancreatic β‐cells. (a) Gene expressions that changed significantly in human TSPAN7 KD pseudo‐islet. Volcano plot shows that genes involved in the pathways of secretion (yellow area) and regulation of transport (green area) were down‐regulated; markers of the apoptotic pathway (blue) were up‐regulated. TSPAN7 KD was demonstrated by the marked reduction in TSPAN7 expression (red). (b) Dynamic insulin secretion measured from perifused human pseudo‐islets of TSPAN7 KD (red) or control (Scramble, black) in response to different glucose concentrations and 20 mM K+ as indicated. Inset: area of the curve (AUC) of insulin secretion stimulated by 20 mM K+. N = 3 for control and TSPAN7 KD, *p < 0.05 versus insulin secretion at 1 mM glucose in the same group of pseudo‐islets; †\\np < 0.05 versus control (Scramble), Student's t‐test. (c) Examples of exocytosis of TSPAN7 KD EndoC‐βH1 (red) or control EndoC‐βH1 (Scramble, black) triggered by depolarisation from −70 to 0 mV with progressively increasing duration (as indicated). (d) Summary of control (black) and exocytosis (normalised to cell size, fF/pF) in relationship to pulse durations. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (e) The relationship between depolarisation‐triggered charge influx and pulse durations of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. *p < 0.05 versus control, student's t‐test. (f) The relationship between charges and exocytosis of control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Corresponding durations of the pulses are as labelled. N = 10 cells for control and 11 cells for TSPAN7 KD EndoC‐βH1 cells. (g) Insulin secretion measured from batch‐incubated control (black) and TSPAN7 KD (red) EndoC‐βH1 cells in response to 1 and 20 mM glucose. Data are the summary of three independent biological repeats in technical triplicate. *p < 0.05 and ***p < 0.001 versus control insulin secretion at 1 mM glucose and ††\\np < 0.01 versus control insulin secretion at 20 mM glucose; two‐way ANOVA and Tuckey post hoc test. (h) Insulin content measured from control (black) and TSPAN7 KD (red) EndoC‐βH1 cells. Data are the summary of three independent biological repeats in technical triplicate. All data are presented as mean value ± SEM.\\nThe insulin secretion experiments suggested that TSPAN7 may play a role in human β‐cell exocytosis. To test this, we performed capacitance measurements using EndoC‐βH1, a human β‐cell line recapturing functional and genetic features of primary human β‐cells,\\n18\\n transduced with Ad‐ShTSPAN7 or Ad‐Scramble virus. Similar to that observed in mouse Tspan7y/− β‐cells, exocytosis triggered by depolarisation was comparable between TSPAN7‐KD EndoC‐βH1 and control cells (Figure 5c,d). However, depolarisation triggered larger influx of charges (p < 0.05) in TSPAN7‐KD EndoC‐βH1 cells than that of the control (Ad‐Scramble) (Figure 5e). Importantly, the exocytotic Ca2+‐sensitivity of TSPAN7‐KD EndoC‐βH1 was reduced (same QCa2+ could only evoke exocytosis that was ~50% of the control, Figure 5f). This correlated with a significantly reduced GSIS in TSPAN7‐KD EndoC‐βH1 (p < 0.01, Figure 5g) without affecting insulin content (Figure 5h). We measured the changes in gene expression of VGCC and synaptotagmins in TSPAN7‐KD EndoC‐βH1. Concomitantly to TSPAN7 KD (p = 5.51 E−4), CACNA1A was up‐regulated and SYT5 and SYT7 were down‐regulated (−20%, Figure 6).\\nChanges in VGCC‐ and synaptotagmin‐encoding genes in EndoC‐βH1 following TSPAN7 KD. The expression of genes encoding TSPAN7 (TSPAN7), VGCCs (CACNA1A, CACNA1C and CACNA1D) and synaptotagmins (SYT4, SYT5 and SYT7) measured from control (transduced with Ad‐Scramble, Scramble, black) and TSPAN7 KD (red) EndoC‐βH1 cells. The levels of expression are normalised to that in control cells (as 100%). Error bars of the control cells condition reflect the variation of the control across the biological repeats. Data are summary of measurements made in two to four independent biological replicates. ***p < 0.0001 versus control, Student's t‐test.\",\n \"Autoantigens implicated in type 1 diabetes have been shown to play a role in the regulation of insulin secretion.\\n25\\n Tspan7, a recently identified T1D autoantigen, can interact with both intracellular and membrane proteins, regulating diverse cellular processes, in a variety of cell types.\\n5\\n While Tspan7 expression in β‐cells is well established,\\n8\\n, \\n13\\n, \\n22\\n its role in insulin secretion is only partially understood. Here, we show that, in addition to its role in regulating VGCCs,\\n13\\n adequate expression of Tspan7 is required to maintain Ca2+ sensitivity of β‐cell exocytosis, which is associated with appropriate expression of synaptotagmin 7. This correlation is found in both mouse and human β‐cells but exerts a more significant role in human β‐cell function.\\nThe role of Tspan7 in regulating cellular functions has been studied in many different cell types\\n5\\n, \\n6\\n, \\n9\\n, \\n10\\n, \\n12\\n, \\n23\\n and this has been recently extended into the study of pancreatic β‐cells,\\n13\\n where it was found to reduce L‐type VGCC activity. Therefore, reducing Tspan7 expression in β‐cells would be expected to lead to higher Ca2+ influx, improving glucose tolerance through enhanced GSIS (a Ca2+‐dependent process). Indeed, in Tspan7y/− β‐cells, depolarisation‐triggered Ca2+ current was augmented (Figure 4a; Figure S3a) and this may contribute to the trend of increased action potential peak, decreased inter‐spike plateau potential and lower firing frequency at high glucose (Figure 3b–d), similar to the effect of an L‐type VGCC agonist BAYK 8644.\\n26\\n However, this increment (~ + 3pA/pF at 0 mV) did not significantly change β‐cell action potential (Figure S1) and thus, may not be sufficient to impact on insulin secretion. This is confirmed by the normal GSIS observed in Tspan7y/− mice, both in vivo (Figure S1) and in vitro (Figure 4f). Similarly, GSIS was not significantly altered in human pseudo‐islets transduced with TSPAN7 shRNA. This differs from the previous report by Dickerson et al.\\n13\\n and may result from the different experimental paradigms used across the two studies (i.e., 2D single cell culture or 3D pseudo‐islets). Pseudo‐islets, formed by re‐aggregating dispersed islet cells, is an approximation of the primary intact islet, preserving a more physiological environment that includes intra‐islet paracrine signalling. Pseudo‐islet function is comparable with primary human islets\\n17\\n and allows efficient genetic manipulation using transduction, as highlighted in the present study (Figure 5a).\\nThe stimulus‐secretion coupling of β‐cells involves exocytosis that depends on Ca2+ influx through the L‐type VGCC.\\n14\\n However, the observation that depolarisation‐triggered insulin secretion was unaffected in Tspan7y/− islets (Figure 4f; Figure S4) and slightly reduced in TSPAN7 KD human pseudo‐islets (Figure 5b) suggests reduced Tspan7/TSPAN7 may also affect exocytotic machinery. By correlating the transmembrane Ca2+ influx with high‐resolution exocytosis measurements, it is clear that exocytosis Ca2+ sensitivity is impaired when the level of tetraspanin‐7 in β‐cells is low. This effect could be in part attributed to reduced expression of Syt7, the major β‐cell exocytosis Ca2+ sensor.\\n27\\n Down‐regulation of the Syt7 following tetraspanin‐7 ablation/knockdown may be due to: (1) a coordinated expressions of these genes, or (2) Ca2+‐dependent changes in gene expression (secondary to augment VGCC activity, reviewed in\\n28\\n). Either possibility suggests that expression of tetraspanin‐7 may be required for appropriate assembly of exocytotic machinery in β‐cells. It is therefore likely that TSPAN7 plays a more significant role in insulin secretion in human β‐cells, in which exocytosis is less dependent on L‐type VGCC (more P/Q‐type VGCC dependent\\n29\\n). This notion is consistent with the observation that insulin secretion induced by 20 mM K+ and GSIS was reduced in TSPAN7 KD pseudo‐islets and EndoC‐βH1 cells respectively (Figure 5b,g). It is interesting to note that the impact of TSPAN7 KD on pseudo‐islets GSIS is less pronounced. This highlights the difference between the primary islets or reconstituted organoids and human cell lines. Whereas the primary human islets (and the pseudo‐islets) are comprised of different type of endocrine cells, EndoC‐βH1 cells are a single β‐cell population, lacking of the local paracrine cross‐talks that are important for appropriate GSIS. For example, it has been reported recently that α‐cells facilitate GSIS through glucagon/GLP‐1‐dependent signalling pathways.\\n30\\n Therefore, it is possible that defects in exocytosis (as the consequence of TSPAN7 KD) could exert a stronger effect in EndoC‐βH1 cells. It will be interesting to test whether a significantly reduced GSIS can be also observed in TSPAN7 KD pseudo‐islet composed by pure human primary β‐cells.\\n\\nTSPAN7 knockdown related transcription change is not restricted to exocytotic pathways. There is a small but significant up‐regulation of genes involved in apoptosis. It is possible this is due to augmented VGCC activity and secondary to elevated cytosolic Ca2+, which is pro‐apoptotic (reviewed in24). However, the degree of up‐regulation in apoptosis‐related genes may not be sufficient to reduce β‐cell mass in either acute or chronic TSPAN7 loss of function, as evidenced by the normal islet insulin content in shTSPAN7 transduced EndoC‐βH1 cells and in aged Tspan7y/− mice.\\nIn summary, in addition to its inhibitory role on β‐cell VGCCs, we found that Tspan7 is required for appropriate expression of genes involved in cell survival and exocytosis. The non‐apparent metabolic phenotype and normal GSIS found in Tspan7y/− mice may be the result of a combined effect of enhanced VGCC activity and reduced sensitivity of exocytosis to Ca2+ in β‐cells.\",\n \"The authors do not have any relevant conflict of interest to disclose.\",\n \"\\nFigure S1.\\n\\nClick here for additional data file.\\n\\nFigure S2.\\n\\nClick here for additional data file.\\n\\nFigure S3.\\n\\nClick here for additional data file.\\n\\nFigure S4.\\n\\nClick here for additional data file.\\n\\nFigure S5.\\n\\nClick here for additional data file.\\n\\nData S1.\\n\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,"methods",null,null,null,null,null,null,null,null,null,null,"results",null,null,null,null,"discussion","COI-statement","supplementary-material"],"string":"[\n null,\n null,\n null,\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n \"discussion\",\n \"COI-statement\",\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["β‐cell","exocytosis","insulin","synaptotagmin","tetraspanin‐7"],"string":"[\n \"β‐cell\",\n \"exocytosis\",\n \"insulin\",\n \"synaptotagmin\",\n \"tetraspanin‐7\"\n]"}}},{"rowIdx":363,"cells":{"title":{"kind":"string","value":"Colorectal Cancer Survival Trends in the United States From 1992 to 2018 Differ Among Persons From Five Racial and Ethnic Groups According to Stage at Diagnosis: A SEER-Based Study."},"pmid":{"kind":"string","value":"36264283"},"background_abstract":{"kind":"string","value":"Survival following colorectal cancer (CRC) has improved in the US since 1975, but there is limited information on stage-specific survival trends among racial and ethnic subgroups."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"We performed a retrospective cohort study of individuals diagnosed with CRC using the 1992-2018 Surveillance, Epidemiology and End Results (SEER) database. We estimated and compared time trends in 1- and 5-year survival for CRC stage by race/ethnicity."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Data from 399 220 individuals diagnosed with CRC were available. There were significant differences in stage-specific 1-year survival trends by race and ethnicity. Differences were most notable for distant stage CRC: survival probabilities increased most consistently for non-Hispanic American Indian/Alaska Native (AIAN) and Black (NHB) persons, but their trend lines were lower than those of Hispanic, and non-Hispanic Asian/Pacific Islander (API) and White (NHW) persons, whose initially greater gains appear to be slowing. Although the data do not support significant racial/ethnic differences in 5-year CRC survival trends by stage, AIAN and NHB persons have the lowest average survival probabilities for multiple CRC stages, and no racial/ethnic group has 5-year survival probabilities above 20% for distant-stage CRC."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Although there has been an overall improvement in adjusted CRC-specific survival probabilities since 1992, AIAN and NHB persons continue to experience worse prognosis than those of other races/ethnicities. This highlights the importance of reinvigorating efforts to understand the causes of mortality in CRC, including those which may differ according to an individual's race or ethnicity."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["United States","Humans","Ethnicity","Retrospective Studies","Racial Groups","Hispanic or Latino","Colorectal Neoplasms"],"string":"[\n \"United States\",\n \"Humans\",\n \"Ethnicity\",\n \"Retrospective Studies\",\n \"Racial Groups\",\n \"Hispanic or Latino\",\n \"Colorectal Neoplasms\"\n]"},"pmcid":{"kind":"string","value":"9597478"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer deaths in the United States.1,2 It has been estimated that 151 030 people will be newly diagnosed with CRC and 52 580 people with die from CRC in 2022.2,3 Although CRC mortality rates have decreased significantly since 1975,1-5 available reports suggest that these declines vary substantially by the stage of the disease, and that disparities in mortality rates for racial and ethnically minoritized groups persist.4-15 For example, according to a report by the American Cancer Society, the 2-year relative survival rate for distant-stage CRC increased from 21% in mid-1990’s to 37% for those diagnosed during 2009-2015, with commensurate improvements for regional- and localized-stage CRC.2 This same report asserts that CRC mortality rates are highest for persons who are non-Hispanic Black, followed by those who are American Indian or Alaska Native, and lowest for those who are Asian or Pacific Islander.2\nSeveral studies have compared trends in CRC mortality by the stage of the disease and race but have done so under broad classifications as follows: White and Black;4,7-10 White, Black, and Hispanic;11,16,17 White, Black, Asian, Hispanic, and other;18 or White, Black, and other.19 In particular, limited information is available on CRC survival trends in persons who are American Indian or Alaska Native.20-22 A 2010 study by Edwards et al,5 compared CRC mortality trends in persons belonging to five racial and ethnic groups, but more than 10 years have passed since that report.\nTo the best of our knowledge, no study has simultaneously compared recent trends in CRC-specific mortality by stage of diagnosis for individuals that are non-Hispanic White, non-Hispanic Black, Hispanic, non-Hispanic Asian or Pacific Islander, and non-Hispanic American Indian or Alaska Native. To address this gap, we undertook an analysis of population-based CRC cause-specific survival data to understand the trends in 1- and 5-year CRC cause-specific survival probabilities by stage of diagnosis within five racial and ethnic groups for individuals who received a CRC diagnosis between 1992 and 2018. Understanding differences in cause-specific survival, both over time and across CRC stages, is critical for further investigations into the social, structural, and political determinants that contribute to the disparities noted in CRC outcomes among individuals from distinct racial and ethnic groups. Through our efforts, we aim to provide new impetus to refocus efforts on improving CRC detection and treatment among racial and ethnically minoritized populations."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Study Design and Data Source [SUBSECTION] This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\nThis study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\n[SUBTITLE] Study Population and Variables [SUBSECTION] We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\nWe used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\n[SUBTITLE] Statistical Analysis [SUBSECTION] We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\nWe summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"This study was based on the 1975-2019 SEER Research Plus Data (November 2021 Submission), which contains information on 425 520 CRC diagnoses from 1992 through 2018. Excluding those individuals without a declared race or ethnicity, and with unknown stage at diagnosis, removed 1372 (.3%) and 24 928 (5.9%) individuals, respectively. This report is based on analysed data from 399 220 individual CRC diagnoses with known race/ethnicity and stage at diagnosis from 1992 through 2018, or 93.8% of all CRC diagnoses from this time period. Table 1 contains tabulations of the numbers of individuals according to levels of the variables of interest, overall and within race and ethnic groups. There was at least 1 year of follow-up for 318 443 (79.8%) persons with CRC eligible at baseline.Table 1.Characteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.AIANAPIHispanicNHBNHWTotalN%n%n%n%n%n%Total39311.045 86911.541 94710.534 1648.6273 30968.5399 220100.0SexMale197950.324 40053.222 21653.016 67448.8139 89751.2205 16651.4Female195249.721 46946.819 73147.017 49051.2133 41248.8194 05448.6Age<40 years2145.415723.424155.812173.659992.211 4172.940-49 years45911.740708.9455310.9340110.016 4666.028 9497.350-59 years91823.4924120.1930822.2781322.941 01815.068 29817.160-69 years104226.511 40424.910 65625.4916126.860 83922.393 10223.370-79 years86522.011 40724.9913021.8777222.778 10828.6107 28226.980+ years43311.0817517.8588514.0480014.070 87925.990 17222.6Summary stageLocalized152038.719 18241.816 57539.513 49139.5115 25442.2166 02241.6Regional149137.917 83338.916 15938.512 00135.1103 89638.0151 38037.9Distant92023.4885419.3921322.0867225.454 15919.881 81820.5Year of diagnosis1992-953488.9458310.036238.6407411.943 15115.855 77914.01996-9945411.6558312.2451410.8449113.244 92516.559 96715.02000-0348112.2671014.6527712.6495914.544 34316.261 77015.42004-0758714.9711615.5636815.2543915.940 99215.060 50215.22008-1165716.7781117.0722017.2560716.438 19213.959 48714.92012-1577819.8789717.2799719.0540615.835 53613.057 61414.42016-1862615.9616913.4694816.6418812.326 1709.544 10111.0GradeI3759.534977.6433510.331539.223 9368.835 2968.8II226057.527 75760.523 47756.019 37556.7155 56256.9228 43157.2III44811.4626913.7575313.7434412.746 96117.263 77516.0IV561.44451.05471.34151.243941.658571.5Unknown79220.1790117.2783518.7687720.142 45615.565 86116.5Rural-urban continuum codeMetro counties150038.240 55788.438 81492.533 18697.1228 17983.5342 23685.7Non-metro counties (metro-adjacent)3839.71120.213013.15401.623 5848.625 9206.5Non-metro counties (not metro-adjacent)39610.116543.616914.04101.220 4267.524 5776.2Alaska or Hawaii164441.800.050.000.000.016490.4Unknown80.235467.71360.3280.111200.448381.2\nCharacteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.\n[SUBTITLE] One- and Five-Year Survival: All Persons With CRC [SUBSECTION] Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\nAdjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\n[SUBTITLE] One-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\nEstimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\n[SUBTITLE] Five-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\nEstimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01)."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Study Design and Data Source","Study Population and Variables","Statistical Analysis","One- and Five-Year Survival: All Persons With CRC","One-Year Survival by Race and Ethnicity","Five-Year Survival by Race and Ethnicity"],"string":"[\n \"Study Design and Data Source\",\n \"Study Population and Variables\",\n \"Statistical Analysis\",\n \"One- and Five-Year Survival: All Persons With CRC\",\n \"One-Year Survival by Race and Ethnicity\",\n \"Five-Year Survival by Race and Ethnicity\"\n]"},"other_sections_texts":{"kind":"list like","value":["This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.","We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.","We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.","Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).","Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.","Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01)."],"string":"[\n \"This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\",\n \"We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\",\n \"We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\",\n \"Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\",\n \"Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\",\n \"Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Study Design and Data Source","Study Population and Variables","Statistical Analysis","Results","One- and Five-Year Survival: All Persons With CRC","One-Year Survival by Race and Ethnicity","Five-Year Survival by Race and Ethnicity","Discussion"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Study Design and Data Source\",\n \"Study Population and Variables\",\n \"Statistical Analysis\",\n \"Results\",\n \"One- and Five-Year Survival: All Persons With CRC\",\n \"One-Year Survival by Race and Ethnicity\",\n \"Five-Year Survival by Race and Ethnicity\",\n \"Discussion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer deaths in the United States.1,2 It has been estimated that 151 030 people will be newly diagnosed with CRC and 52 580 people with die from CRC in 2022.2,3 Although CRC mortality rates have decreased significantly since 1975,1-5 available reports suggest that these declines vary substantially by the stage of the disease, and that disparities in mortality rates for racial and ethnically minoritized groups persist.4-15 For example, according to a report by the American Cancer Society, the 2-year relative survival rate for distant-stage CRC increased from 21% in mid-1990’s to 37% for those diagnosed during 2009-2015, with commensurate improvements for regional- and localized-stage CRC.2 This same report asserts that CRC mortality rates are highest for persons who are non-Hispanic Black, followed by those who are American Indian or Alaska Native, and lowest for those who are Asian or Pacific Islander.2\nSeveral studies have compared trends in CRC mortality by the stage of the disease and race but have done so under broad classifications as follows: White and Black;4,7-10 White, Black, and Hispanic;11,16,17 White, Black, Asian, Hispanic, and other;18 or White, Black, and other.19 In particular, limited information is available on CRC survival trends in persons who are American Indian or Alaska Native.20-22 A 2010 study by Edwards et al,5 compared CRC mortality trends in persons belonging to five racial and ethnic groups, but more than 10 years have passed since that report.\nTo the best of our knowledge, no study has simultaneously compared recent trends in CRC-specific mortality by stage of diagnosis for individuals that are non-Hispanic White, non-Hispanic Black, Hispanic, non-Hispanic Asian or Pacific Islander, and non-Hispanic American Indian or Alaska Native. To address this gap, we undertook an analysis of population-based CRC cause-specific survival data to understand the trends in 1- and 5-year CRC cause-specific survival probabilities by stage of diagnosis within five racial and ethnic groups for individuals who received a CRC diagnosis between 1992 and 2018. Understanding differences in cause-specific survival, both over time and across CRC stages, is critical for further investigations into the social, structural, and political determinants that contribute to the disparities noted in CRC outcomes among individuals from distinct racial and ethnic groups. Through our efforts, we aim to provide new impetus to refocus efforts on improving CRC detection and treatment among racial and ethnically minoritized populations.","[SUBTITLE] Study Design and Data Source [SUBSECTION] This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\nThis study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\n[SUBTITLE] Study Population and Variables [SUBSECTION] We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\nWe used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\n[SUBTITLE] Statistical Analysis [SUBSECTION] We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\nWe summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.","This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.","We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.","We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.","This study was based on the 1975-2019 SEER Research Plus Data (November 2021 Submission), which contains information on 425 520 CRC diagnoses from 1992 through 2018. Excluding those individuals without a declared race or ethnicity, and with unknown stage at diagnosis, removed 1372 (.3%) and 24 928 (5.9%) individuals, respectively. This report is based on analysed data from 399 220 individual CRC diagnoses with known race/ethnicity and stage at diagnosis from 1992 through 2018, or 93.8% of all CRC diagnoses from this time period. Table 1 contains tabulations of the numbers of individuals according to levels of the variables of interest, overall and within race and ethnic groups. There was at least 1 year of follow-up for 318 443 (79.8%) persons with CRC eligible at baseline.Table 1.Characteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.AIANAPIHispanicNHBNHWTotalN%n%n%n%n%n%Total39311.045 86911.541 94710.534 1648.6273 30968.5399 220100.0SexMale197950.324 40053.222 21653.016 67448.8139 89751.2205 16651.4Female195249.721 46946.819 73147.017 49051.2133 41248.8194 05448.6Age<40 years2145.415723.424155.812173.659992.211 4172.940-49 years45911.740708.9455310.9340110.016 4666.028 9497.350-59 years91823.4924120.1930822.2781322.941 01815.068 29817.160-69 years104226.511 40424.910 65625.4916126.860 83922.393 10223.370-79 years86522.011 40724.9913021.8777222.778 10828.6107 28226.980+ years43311.0817517.8588514.0480014.070 87925.990 17222.6Summary stageLocalized152038.719 18241.816 57539.513 49139.5115 25442.2166 02241.6Regional149137.917 83338.916 15938.512 00135.1103 89638.0151 38037.9Distant92023.4885419.3921322.0867225.454 15919.881 81820.5Year of diagnosis1992-953488.9458310.036238.6407411.943 15115.855 77914.01996-9945411.6558312.2451410.8449113.244 92516.559 96715.02000-0348112.2671014.6527712.6495914.544 34316.261 77015.42004-0758714.9711615.5636815.2543915.940 99215.060 50215.22008-1165716.7781117.0722017.2560716.438 19213.959 48714.92012-1577819.8789717.2799719.0540615.835 53613.057 61414.42016-1862615.9616913.4694816.6418812.326 1709.544 10111.0GradeI3759.534977.6433510.331539.223 9368.835 2968.8II226057.527 75760.523 47756.019 37556.7155 56256.9228 43157.2III44811.4626913.7575313.7434412.746 96117.263 77516.0IV561.44451.05471.34151.243941.658571.5Unknown79220.1790117.2783518.7687720.142 45615.565 86116.5Rural-urban continuum codeMetro counties150038.240 55788.438 81492.533 18697.1228 17983.5342 23685.7Non-metro counties (metro-adjacent)3839.71120.213013.15401.623 5848.625 9206.5Non-metro counties (not metro-adjacent)39610.116543.616914.04101.220 4267.524 5776.2Alaska or Hawaii164441.800.050.000.000.016490.4Unknown80.235467.71360.3280.111200.448381.2\nCharacteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.\n[SUBTITLE] One- and Five-Year Survival: All Persons With CRC [SUBSECTION] Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\nAdjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\n[SUBTITLE] One-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\nEstimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\n[SUBTITLE] Five-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\nEstimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).","Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).","Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.","Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).","The purpose of this study was to estimate and compare race- and ethnicity-specific 1- and 5-year CRC cause-specific survival trends, within the context of the stage at which the cancer was first detected. We utilized data from the SEER registries for persons diagnosed with CRC from 1992 to 2018 and estimated 1- and 5-year survival probabilities within groups of persons with CRC by race and ethnicity, stage of disease, and year of diagnosis. Study findings contribute to existing knowledge gaps regarding trends in survival by including estimates for an often-overlooked population: American Indian or Alaska Natives.\nThere are significant differences in the degree to which improvements in 1-year survival probabilities following stage-specific CRC are experienced according to a person’s race and ethnicity. The estimates that we report specifically for AIAN and API subgroups contribute towards the limited literature focused on identifying survival trend differences for these racially minoritized populations. Study findings suggest that AIAN individuals have the lowest 1-year survival probabilities compared to those of other races. Lower 1-year CRC-specific survival probabilities are also seen among NHB persons. In particular, these persons have the lowest average survival probabilities following the diagnosis of distant-stage CRC. Although differences among racial and ethnic groups have moderated somewhat over the study period, these disparities do not appear to have resolved completely.\nThere is established evidence that CRC screening can prevent or detect CRC early.27,28 For screening to be effective in improving outcomes, timely follow up of any abnormal test is necessary. Structural barriers (eg, lack of insurance or social support, racism and discrimination) to obtaining appropriate primary care services may play an important role in reducing the possibility for racially minoritized populations to receive guideline-compliant screening services.5,29-31 The fact that some studies conducted within the Veterans’ Health Administration found no differences between White and Black persons in diagnostic follow up testing,32,33 suggests that access to appropriate structures and services may play an important role in appropriate post-screening follow-up for minoritized populations. As expected, the greatest differences in survival probabilities are apparent among individuals diagnosed at different stages of CRC. Although these patterns are largely similar among persons from different race and ethnic groups disparities in survival persist. Enhanced follow-up of abnormal results, may help overcome at least some of the persistent disparities in CRC survival probabilities among persons from different race and ethnic groups.\nOther differences in access to and utilization of quality health care may also contribute to the observed differences of survival trends by race.34-36 Some studies have suggested NHB persons are less likely to receive surgical treatment and adjuvant chemotherapy.35-38 Social, structural and political determinants also contribute towards the likelihood of NHB and AIAN persons to be diagnosed with advanced stage CRC compared to NHW persons.39,40 This may further contribute to the lower survival probabilities observed in NHB persons with distant stage CRC.41,42 Racial and ethnic minoritized populations also experience disparities in terms of post-treatment surveillance and distinct baseline comorbidities, which further contribute to lower survival rates.34,35 Several studies continue to highlight that transportation barriers, cultural beliefs, fear and stigma about screening, and concerns about privacy issues are contributing factors to survival outcomes43-47 for persons with CRC.\nLifestyle and biological factors may also play a role in CRC risk and outcomes. For instance, lifestyle factors such as obesity, smoking, alcohol consumption, and physical inactivity have a higher prevalence among Black populations.48 Also, genetic mutations and microsatellite instability can differ among racially and ethnically minoritized populations. All of these things can affect CRC development and prognosis,49-54 and may play into the disparities noted in this work.\nThere are several potential limitations in our analysis. First, we must acknowledge that this is a retrospective analysis of data from the SEER registries. However, SEER is a population-based resource which provides information about the most critical factors of interest, and we tried to control for these. Second, we cannot fully exclude the possibility that regional variations may contribute to the observed CRC survival probability estimates that average across SEER registries. Third, we chose to report on cause-specific survival, and this may be influenced by different practices in cause of death ascertainment. However, in sensitivity analyses performed using relative survival estimates we found that overall findings were largely concordant. This is similar to a report of data from Canada where differences in survival probabilities were noted, and that differences between First Nations and non-aboriginal persons were somewhat smaller for cause-specific survival when compared to relative survival estimates.55 Fourth, SEER data have relatively little information on comorbidities, access to care, and insurance status. Fifth, racial and ethnic classifications in medical records may reflect misclassification,56 which may bias our estimates. Finally, SEER registries represent a subset of AIAN persons spanning the United States, but do not capture data from some regions with large AIAN populations, including those in Oklahoma, Arizona, or the Northern Plains and Great Lakes.57 Even with these limitations, we are able to provide new data concerning trends in CRC survival over time for multiple racial and ethnic groups in the United States, according to their stage at diagnosis.\nIn conclusion, we present trends in 1- and 5-year CRC cause-specific survival probabilities for persons of five racial and ethnic groups. These estimates of, and trends in, survival probabilities for groups of minoritized races and ethnicities may enable the development of a more complete picture of CRC prognosis. We have identified significant differences in the race- and ethnic-specific trends in 1-year stage-specific CRC survival probabilities. In particular, AIAN persons have historically experienced poorer CRC prognosis, as have NHB individuals. Although these disparities appear to be lessening somewhat, the current differences in survival probabilities continue to call for further work in order to erase them completely. Trends in 1-year survival probabilities are significantly different among those diagnosed at different stages of CRC. This, coupled with the finding that the distribution of stages at diagnosis appears to differ among persons of different racial and ethnic groups, suggests that prioritizing specific stages of disease at diagnosis may provide 1 avenue that may help us to overcome race- and ethnic-specific disparities in CRC-specific survival probability. This also argues for a potential need to improve screening for CRC, as this may influence the distribution of stages at which it is diagnosed. Future research should strive to capture cancer incidence and survival information from all key racial and ethnic subgroups. Future efforts should also incorporate and evaluate multi-level interventions at the individual, structural, and policy levels to address the persistent disparities in CRC survival."],"string":"[\n \"Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer deaths in the United States.1,2 It has been estimated that 151 030 people will be newly diagnosed with CRC and 52 580 people with die from CRC in 2022.2,3 Although CRC mortality rates have decreased significantly since 1975,1-5 available reports suggest that these declines vary substantially by the stage of the disease, and that disparities in mortality rates for racial and ethnically minoritized groups persist.4-15 For example, according to a report by the American Cancer Society, the 2-year relative survival rate for distant-stage CRC increased from 21% in mid-1990’s to 37% for those diagnosed during 2009-2015, with commensurate improvements for regional- and localized-stage CRC.2 This same report asserts that CRC mortality rates are highest for persons who are non-Hispanic Black, followed by those who are American Indian or Alaska Native, and lowest for those who are Asian or Pacific Islander.2\\nSeveral studies have compared trends in CRC mortality by the stage of the disease and race but have done so under broad classifications as follows: White and Black;4,7-10 White, Black, and Hispanic;11,16,17 White, Black, Asian, Hispanic, and other;18 or White, Black, and other.19 In particular, limited information is available on CRC survival trends in persons who are American Indian or Alaska Native.20-22 A 2010 study by Edwards et al,5 compared CRC mortality trends in persons belonging to five racial and ethnic groups, but more than 10 years have passed since that report.\\nTo the best of our knowledge, no study has simultaneously compared recent trends in CRC-specific mortality by stage of diagnosis for individuals that are non-Hispanic White, non-Hispanic Black, Hispanic, non-Hispanic Asian or Pacific Islander, and non-Hispanic American Indian or Alaska Native. To address this gap, we undertook an analysis of population-based CRC cause-specific survival data to understand the trends in 1- and 5-year CRC cause-specific survival probabilities by stage of diagnosis within five racial and ethnic groups for individuals who received a CRC diagnosis between 1992 and 2018. Understanding differences in cause-specific survival, both over time and across CRC stages, is critical for further investigations into the social, structural, and political determinants that contribute to the disparities noted in CRC outcomes among individuals from distinct racial and ethnic groups. Through our efforts, we aim to provide new impetus to refocus efforts on improving CRC detection and treatment among racial and ethnically minoritized populations.\",\n \"[SUBTITLE] Study Design and Data Source [SUBSECTION] This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\\nThis study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\\n[SUBTITLE] Study Population and Variables [SUBSECTION] We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\\nWe used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\\n[SUBTITLE] Statistical Analysis [SUBSECTION] We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\\nWe summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\",\n \"This study was based on a retrospective cohort of CRC patients, ascertained on a population level, with data captured as part of the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program. The SEER program collects cancer incidence and survival data from population-based cancer registries representing approximately 35% of the U.S. population.23 Data from the 12 SEER registries for the period 1992 to 2018 were used in this analysis. The 12 SEER registries cover Alaska, Connecticut, Atlanta, rural Georgia, San Francisco-Oakland, San Jose-Monterey, Hawaii, Iowa, Los Angeles, New Mexico, Seattle-Puget Sound and Utah.24 This project, Study ID: 21-102, was reviewed by the University of New Mexico Health Sciences Institutional Review Board, and was granted exemption on 31 March 2021 according to Category 4: Secondary research on data or specimens (no consent required). This report conforms to RECORD guidelines for SEER-based studies.25 This study was supported by funding from the National Cancer Institute; the funders played no other role in this work. Researchers desiring additional details about the data and programs used to carry out this work may obtain programming code from the corresponding author.\",\n \"We used SEER*Stat software (version 8.4.0)26 as the data source for this study. We used the “Incidence––SEER Research Plus Data, 12 Registries, Nov 2021 Sub (1992-2019)” database. We included a consecutive series of all individuals who received their first primary CRC diagnosis with malignant behavior from 1992 through 2018. We excluded “All Death Certificate Only and Autopsy Only” and “Alive with No Survival Time” via checkboxes in SEER*Stat. We also excluded CRC cases without a record for summary stage at diagnosis, as well as those with unknown age at diagnosis. We included all CRC cases that originated from the cecum, ascending colon, hepatic flexure of colon, transverse colon, splenic flexure of colon, descending colon, sigmoid colon, overlapping lesion of colon, colon not otherwise specified, rectosigmoid junction, and rectum. We classified CRC cases as localized, regional and distant stage using SEER’s “Combined Summary Stage (2004+)” classifications, supplemented with values from the “Historic Stage A (1973-2015)” variable when necessary. In-situ cancers and cases that were un-staged/unknown were excluded. These rules, as applied in SEER*Stat 8.0.4, defined the number of individuals included in this study.\\nThe SEER program works closely with providers of cancer care in their population-based catchment areas to collect patient- and cancer-specific information. Because of its population-based nature, it is typically necessary to rely on medical records as the source of the demographic and other data. SEER-defined race and ethnic categories were utilized and labeled as follows: Non-Hispanic White (NHW), Non-Hispanic Black (NHB), Hispanic, Non-Hispanic American Indian/Alaska Native (AIAN), and Non-Hispanic Asian or Pacific Islander (API). Additional variables were extracted from SEER, and estimates of CRC-specific survival probabilities were obtained for combinations of these variables in the primary analyses: sex as assigned at birth (male or female), age at diagnosis (categorized into decades), year of diagnosis, grade (grades I-IV, or unknown), and Rural-Urban Continuum Code in 2003 (RUCC, coded as: Metro Counties, Non-Metro Counties [Metro-Adjacent], Non-Metro Counties [not Metro-Adjacent], and Alaska or Hawaii, or unknown).\\nThe outcome of interest in this work was cause-specific survival for CRC, estimated at 1- and 5-years post diagnosis. We used SEER*Stat software (version 8.4.0)26 to calculate 1- and 5-year cause-specific survival probabilities of CRC using the “Incidence – SEER Research Plus Data, 12 Registries, Nov 2021 (1992-2019)” database while relying on the rules implemented in SEER*Stat for loss to follow-up. We obtained cause-specific survival probabilities and their standard errors within combinations of the characteristics of interest, ie, year of diagnosis, race and ethnicity, stage, sex, grade, and RUCC. No person-level data were used, nor were the SEER data linked to any other data sources. Using SEER*Stat 8.4.0, we formed tables of cause-specific survival probabilities and their standard errors according to all combinations of the factors of interest, with separate “Pages” defined for combinations of Year of diagnosis, age at diagnosis (in decades), and RUCC groups, and with rows defined by combinations of race/ethnicity, summary stage, grade, and sex. We used text processing approaches to form an analysis-ready data set which enumerated the numbers of individuals with a primary CRC diagnosis, and estimates of cause-specific survival probabilities and standard errors, within categories defined by all of the factors of interest in this study.\",\n \"We summarized the numbers and percentages of individuals with CRC diagnoses within the levels of the variables of interest. We used linear models to perform meta-regressions, weighting by the inverse of the squared standard errors, to estimate the degree to which the combinations of the variables of interest explained differences in the estimated cause-specific survival probabilities. When the SEER-estimated standard errors of survival probabilities were equal to zero, we approximated them with the square root of the variance of the binomial distribution, calculated after adding a value of .5 to both the numerator and denominator counts reported by SEER for the relevant group. We modelled the survival probabilities obtained at 1 year and 5 years post-diagnosis separately. The initial models included all possible interactions among race/ethnicity, stage, and year of diagnosis, while controlling for potential confounding by age at diagnosis (categorized into decades), sex, grade, and RUCC. We simplified these initial models subsequently modelling year-at-diagnosis trends with natural cubic spline basis functions, and selected the degree of smoothing that best explained the race by stage by year of diagnosis trends by minimizing Akaike’s Information Criterion (AIC) while adjusting for the other factors. We obtained smoothed trend estimates within groups defined by race and ethnicity, and by CRC stage at diagnosis, from a separate model for each of the two follow-up time periods. These final models were obtained by removing non-significant interactions in a hierarchical fashion. Analyses were performed using the tools available in SEER*Stat26 and SAS 9.4 (SAS Institute Inc, Cary, NC). Two-sided type I error was set at .05 for tests of significance.\",\n \"This study was based on the 1975-2019 SEER Research Plus Data (November 2021 Submission), which contains information on 425 520 CRC diagnoses from 1992 through 2018. Excluding those individuals without a declared race or ethnicity, and with unknown stage at diagnosis, removed 1372 (.3%) and 24 928 (5.9%) individuals, respectively. This report is based on analysed data from 399 220 individual CRC diagnoses with known race/ethnicity and stage at diagnosis from 1992 through 2018, or 93.8% of all CRC diagnoses from this time period. Table 1 contains tabulations of the numbers of individuals according to levels of the variables of interest, overall and within race and ethnic groups. There was at least 1 year of follow-up for 318 443 (79.8%) persons with CRC eligible at baseline.Table 1.Characteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.AIANAPIHispanicNHBNHWTotalN%n%n%n%n%n%Total39311.045 86911.541 94710.534 1648.6273 30968.5399 220100.0SexMale197950.324 40053.222 21653.016 67448.8139 89751.2205 16651.4Female195249.721 46946.819 73147.017 49051.2133 41248.8194 05448.6Age<40 years2145.415723.424155.812173.659992.211 4172.940-49 years45911.740708.9455310.9340110.016 4666.028 9497.350-59 years91823.4924120.1930822.2781322.941 01815.068 29817.160-69 years104226.511 40424.910 65625.4916126.860 83922.393 10223.370-79 years86522.011 40724.9913021.8777222.778 10828.6107 28226.980+ years43311.0817517.8588514.0480014.070 87925.990 17222.6Summary stageLocalized152038.719 18241.816 57539.513 49139.5115 25442.2166 02241.6Regional149137.917 83338.916 15938.512 00135.1103 89638.0151 38037.9Distant92023.4885419.3921322.0867225.454 15919.881 81820.5Year of diagnosis1992-953488.9458310.036238.6407411.943 15115.855 77914.01996-9945411.6558312.2451410.8449113.244 92516.559 96715.02000-0348112.2671014.6527712.6495914.544 34316.261 77015.42004-0758714.9711615.5636815.2543915.940 99215.060 50215.22008-1165716.7781117.0722017.2560716.438 19213.959 48714.92012-1577819.8789717.2799719.0540615.835 53613.057 61414.42016-1862615.9616913.4694816.6418812.326 1709.544 10111.0GradeI3759.534977.6433510.331539.223 9368.835 2968.8II226057.527 75760.523 47756.019 37556.7155 56256.9228 43157.2III44811.4626913.7575313.7434412.746 96117.263 77516.0IV561.44451.05471.34151.243941.658571.5Unknown79220.1790117.2783518.7687720.142 45615.565 86116.5Rural-urban continuum codeMetro counties150038.240 55788.438 81492.533 18697.1228 17983.5342 23685.7Non-metro counties (metro-adjacent)3839.71120.213013.15401.623 5848.625 9206.5Non-metro counties (not metro-adjacent)39610.116543.616914.04101.220 4267.524 5776.2Alaska or Hawaii164441.800.050.000.000.016490.4Unknown80.235467.71360.3280.111200.448381.2\\nCharacteristics of Patients With Colorectal Cancer From 1992 to 2018, by Racial and Ethnic Groups.\\n[SUBTITLE] One- and Five-Year Survival: All Persons With CRC [SUBSECTION] Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\\nAdjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\\n[SUBTITLE] One-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\\nEstimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\\n[SUBTITLE] Five-Year Survival by Race and Ethnicity [SUBSECTION] Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\\nEstimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\",\n \"Adjusted 1- and 5-year cause-specific survival probabilities following CRC diagnosis by stage and year of diagnosis are shown in Figure 1. For those diagnosed with local stage CRC, the estimate of linear trend in 1-year survival probabilities was negligible, at −.01% ([95% Confidence Interval (CI)] −.03%-.01%) per year, with an average 1-year survival probability of 96.2% (93.8%-98.6%). Those diagnosed with regional stage CRC displayed a small improvement over time in 1-year survival probabilities, .06% (.001%-.12%) per year. Their average 1-year survival probability was 93.0% (90.5%-95.5%). For those diagnosed with distant stage CRC, there was a significant improvement in 1-year survival probabilities over time (P < .001), but there was evidence of departure from a linear trend (P = .04). The slope of the trend line suggested an improvement in survival probability of 1.0% (.8%-1.2%) per year in 2005, with instantaneous slopes suggesting greater gains, by .05% (.01%-.09%), for each year prior to 2005 and slower gains by that same amount for each year following 2005. Their average 1-year survival probability was 52.8% (48.1%-57.5%).Figure 1.Trends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nTrends in 1- and 5-year adjusted cause-specific survival probabilities for individuals diagnosed at different stages of colorectal cancer. Shaded bands reflect 95% prediction intervals for the year-specific survival probabilities.\\nThose diagnosed with local stage CRC experienced improvements in 5-year cause-specific survival, with an average improvement of .16% (.10%-.21%) per year. Their average 5-year survival probability was 90.9% (87.9%-93.8%). Those diagnosed with regional stage CRC displayed significant improvement in 5-year survival probabilities over time (P < .001), but there was evidence of significant departure from a linear trend (P = .004), such that the slope of the trend line showed an improvement of .55% (.41%-.69%) per year in 2003, with instantaneous slopes suggested greater gains, by .08% (.04-.12%), for each year prior to 2003 and slower gains by that same amount for each year following 2003. Their average 5-year survival probability was 70.3% (66.5%-74.0%). For those diagnosed with distant stage CRC, there was a significant but small improvement over time in 5-year survival probabilities, .15% (.5%-.25%) per year. Their average 5-year survival probability was 8.1% (4.4%-11.8%).\",\n \"Estimates of adjusted 1-year cause-specific survival probabilities following diagnosis are shown for all race/ethnicity by stage combinations in Tables 2 and 3. Trends in 1-year survival for those of different races/ethnicities who were diagnosed at different stages of CRC differed significantly (P < .001, see Figure 2). For those diagnosed with localized stage CRC, 1-year survival probabilities were consistently high for individuals of all race and ethnic groups. AIAN persons experienced lower 1-year survival probabilities than those from other race and ethnic groups over the study period. Although due to lack of precision in the estimates, the difference was only significantly different when compared to API persons; average 1-year survival probabilities were 1.5 (.5-2.5) percentage points lower for AIAN than for API persons.Table 2.Unadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN1992100 (84.6-100)91.6 (75.4-100)27.0 (5.6-48.4)72.9 (62.6-83.2)AIAN199696.9 (81.8-100)95.7 (77.9-100)41.8 (23.5-60.1)78.1 (68.2-88.0)AIAN200098.7 (83.6-100)83.2 (65.9-100)43.0 (20.0-66.0)75.0 (64.1-85.8)AIAN200496.7 (85.6-100)96.4 (84.8-100)51.8 (35.8-67.7)81.6 (74.1-89.2)AIAN2008100 (86.4-100)95.5 (83.0-100)53.5 (35.5-71.5)83.0 (74.4-91.6)AIAN2012100 (88.9-100)91.6 (79.1-100)52.7 (38.0-67.4)81.4 (74.0-88.8)AIAN201696.0 (85.5-100)95.9 (85.6-100)46.5 (33.4-59.6)79.5 (72.9-86.0)AIAN2018100 (90.7-100)99.3 (88.9-100)83.5 (67.2-99.8)94.3 (87.1-100)API199299.5 (96.2-100)94.6 (90.8-98.3)46.5 (38.4-54.6)80.2 (77.0-83.3)API199699.4 (97.1-100)95.8 (92.6-98.9)56.5 (48.7-64.3)83.9 (81.0-86.8)API200099.2 (97.0-100)97.9 (95.8-99.9)52.8 (45.0-60.6)83.3 (80.5-86.1)API200499.6 (97.9-100)97.3 (95.1-99.5)68.6 (62.7-74.4)88.5 (86.3-90.6)API200899.7 (98.3-100)98.4 (96.5-100)59.5 (53.9-65.0)85.9 (83.9-87.9)API201299.4 (97.9-100)96.9 (94.7-99.2)55.8 (49.4-62.3)84.1 (81.7-86.4)API201699.1 (97.4-100)98.4 (96.7-100)57.6 (51.8-63.3)85.0 (82.9-87.1)API201899.6 (98.5-100)97.8 (96.0-99.6)69.1 (62.7-75.5)88.8 (86.6-91.1)Hispanic199297.0 (93.0-100)96.7 (93.3-100)55.6 (46.2-64.9)83.1 (79.5-86.7)Hispanic199698.8 (95.8-100)96.1 (92.7-99.5)43.4 (35.7-51.1)79.4 (76.4-82.4)Hispanic200099.2 (96.6-100)96.3 (93.3-99.4)47.0 (39.8-54.2)80.8 (78.1-83.6)Hispanic200498.9 (96.5-100)95.8 (92.9-98.7)57.8 (50.8-64.8)84.2 (81.5-86.8)Hispanic200899.1 (97.3-100)96.2 (93.5-98.8)60.7 (54.4-67.1)85.3 (83.0-87.7)Hispanic201299.2 (97.4-100)95.8 (93.3-98.2)71.6 (66.6-76.7)88.9 (86.9-90.8)Hispanic201698.9 (97.1-100)98.2 (96.6-99.7)69.0 (64.0-74.1)88.7 (86.9-90.6)Hispanic201899.6 (98.6-100)97.8 (96.1-99.5)58.8 (52.5-65.0)85.4 (83.2-87.6)NHB199298.0 (94.7-100)93.7 (89.6-97.8)38.5 (30.0-47.0)76.7 (73.4-80.1)NHB199698.2 (94.7-100)96.3 (93-99.6)30.1 (23.1-37.2)74.9 (72.0-77.7)NHB200097.3 (93.8-100)93.2 (89.4-97.0)34.5 (27.3-41.7)75.0 (72.0-77.9)NHB200498.4 (95.6-100)97.0 (94.3-99.7)37.0 (30.7-43.3)77.5 (75.0-79.9)NHB200898.6 (96.2-100)93.3 (89.4-97.2)51.4 (44.7-58.1)81.1 (78.4-83.8)NHB201299.5 (97.2-100)93.6 (90.0-97.1)68.5 (62.1-74.8)87.2 (84.7-89.7)NHB201699.5 (97.2-100)96.0 (93.0-99.0)53.1 (47.2-58.9)82.9 (80.5-85.2)NHB201899.3 (97.5-100)97.0 (94.2-99.8)57.1 (49.6-64.6)84.5 (81.7-87.2)NHW199298.6 (97.9-99.2)94.6 (93.5-95.7)38.0 (35.0-40.9)77.0 (76.0-78.1)NHW199698.9 (98.3-99.5)96.1 (95.1-97.0)36.7 (33.8-39.5)77.2 (76.2-78.2)NHW200099.0 (98.5-99.6)94.3 (93.2-95.4)40.9 (37.9-43.9)78.1 (77.0-79.1)NHW200498.9 (98.3-99.5)96.1 (95.1-97.1)52.6 (49.7-55.4)82.5 (81.5-83.6)NHW200899.2 (98.6-99.7)97.0 (96.1-98.0)58.6 (55.6-61.7)85.0 (83.9-86.0)NHW201298.9 (98.2-99.6)97.2 (96.2-98.2)55.2 (52.3-58.0)83.8 (82.7-84.8)NHW201699.1 (98.5-99.8)96.9 (95.9-97.9)60.7 (57.9-63.5)85.6 (84.6-86.6)NHW201898.7 (98.0-99.4)97.5 (96.7-98.4)59.0 (55.7-62.3)85.1 (83.9-86.3)Table 3.Adjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199297.5 (83.7-100)89.8 (75.3-100)25.4 (6.2-44.6)70.9 (61.6-80.2)AIAN199695.6 (82.0-100)93.2 (77.3-100)41.9 (25.5-58.3)76.9 (68.0-85.8)AIAN200097.0 (83.5-100)82.5 (67.0-97.9)42.0 (21.4-62.6)73.8 (64.1-83.5)AIAN200495.3 (85.3-100)94.2 (83.8-100)50.4 (36.1-64.7)80.0 (73.2-86.8)AIAN200897.5 (85.3-100)92.4 (81.2-100)51.6 (35.5-67.7)80.5 (72.8-88.2)AIAN201297.5 (87.5-100)89.9 (78.7-100)50.2 (37.0-63.3)79.2 (72.5-85.9)AIAN201693.4 (84.0-100)94.1 (84.8-100)45.0 (33.3-56.8)77.5 (71.6-83.4)AIAN201896.7 (87.8-100)96.9 (87.6-100)64.2 (51.1-77.5)85.9 (79.7-91.7)API199297.5 (94.5-100)92.3 (88.9-95.7)45.8 (38.6-53.1)78.6 (75.7-81.4)API199697.4 (95.2-99.6)93.4 (90.5-96.3)55.6 (48.5-62.6)82.1 (79.4-84.8)API200097.4 (95.3-99.5)95.9 (93.9-97.9)53.1 (46.1-60.1)82.1 (79.6-84.7)API200497.8 (96.2-99.4)94.6 (92.5-96.7)67.1 (61.8-72.4)86.5 (84.5-88.5)API200897.5 (96.0-98.9)95.6 (93.7-97.4)59.4 (54.4-64.4)84.2 (82.2-86.1)API201296.8 (95.3-98.3)94.6 (92.5-96.7)56.1 (50.3-61.9)82.5 (80.3-84.7)API201696.4 (94.8-98.0)95.0 (93.4-96.6)57.8 (52.6-63.0)83.1 (81.1-85.0)API201897.9 (96.5-99.0)96.1 (94.5-97.9)63.8 (58.3-69.3)86.0 (83.9-88.0)Hispanic199294.7 (91.1-98.4)94.4 (91.3-97.5)54.3 (45.9-62.7)81.2 (77.9-84.4)Hispanic199696.8 (93.9-99.6)94.2 (91.0-97.3)42.9 (36.0-49.9)77.9 (75.2-80.7)Hispanic200097.0 (94.6-99.4)93.4 (90.6-96.2)46.6 (40.1-53.0)79.0 (76.4-81.5)Hispanic200497.2 (95.0-99.4)93.3 (90.6-96.0)57.1 (50.8-63.4)82.5 (80.1-85.0)Hispanic200896.2 (94.5-97.9)92.8 (90.4-95.3)59.8 (54.1-65.5)83.0 (80.8-85.2)Hispanic201296.0 (94.3-97.8)92.9 (90.6-95.2)69.3 (64.8-73.9)86.1 (84.2-88.0)Hispanic201694.8 (93.0-96.5)94.1 (92.5-95.6)67.3 (62.8-71.9)85.4 (83.6-87.2)Hispanic201896.8 (95.3-98.3)95.3 (93.7-96.9)63.4 (58.3-68.6)85.2 (83.2-87.2)NHB199295.7 (92.6-98.7)91.1 (87.3-94.8)37.8 (30.2-45.4)74.8 (71.8-77.9)NHB199696.6 (93.4-99.9)93.8 (90.7-96.8)30.8 (24.4-37.1)73.7 (71.1-76.3)NHB200095.6 (92.5-98.8)90.5 (87.1-94.0)35.7 (29.2-42.2)74.0 (71.2-76.7)NHB200495.9 (93.3-98.5)94.0 (91.4-96.5)38.9 (33.2-44.6)76.3 (73.9-78.6)NHB200896.1 (93.9-98.3)90.3 (86.8-93.9)51.2 (45.1-57.2)79.2 (76.7-81.7)NHB201296.6 (94.4-98.7)90.6 (87.4-93.8)66.7 (60.9-72.4)84.6 (82.2-87.0)NHB201696.0 (93.9-98.2)92.3 (89.6-95.1)53.0 (47.7-58.3)80.5 (78.3-82.6)NHB201897.5 (95.6-99.1)94.3 (91.7-97.1)55.2 (49.2-61.3)82.4 (80.0-84.8)NHW199296.8 (95.9-97.7)92.8 (91.6-94.0)39.2 (36.5-42.0)76.3 (75.1-77.4)NHW199696.7 (95.9-97.6)93.1 (92.1-94.2)38.4 (35.8-41.1)76.1 (75.0-77.2)NHW200096.3 (95.5-97.2)92.4 (91.2-93.6)42.4 (39.6-45.1)77.0 (75.9-78.2)NHW200496.5 (95.6-97.3)93.5 (92.4-94.7)53.5 (50.8-56.1)81.2 (80.0-82.3)NHW200896.4 (95.6-97.3)93.8 (92.7-94.9)58.8 (56.0-61.6)83.0 (81.8-84.2)NHW201295.9 (95.0-96.8)93.9 (92.8-95.0)55.7 (53.0-58.3)81.8 (80.7-83.0)NHW201695.2 (94.3-96.0)93.5 (92.4-94.6)60.4 (57.8-63.0)83.0 (81.9-84.2)NHW201896.7 (95.8-97.6)95.0 (93.95-96.1)60.4 (57.6-63.3)84.0 (82.9-85.3)Figure 2.Trends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nUnadjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 1-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 1-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2018.\\nFor those diagnosed with regional stage CRC, trends in adjusted 1-year cause-specific survival probabilities similarly reflect little change over time, with the possible exception of an imprecisely-estimated positive trend over time among AIAN persons. Although there was little evidence of significant trends over time, there was a clear gradient in the average survival probabilities among those from different race and ethnic groups. AIAN persons had the lowest average survival probability (90.2%, 88.2%-92.2), followed by NHB, NHW, Hispanic, and API persons (93.2%, 92.6%-93.8%; 93.3%, 93.1%-93.5%; 94.2%, 93.8%-94.6%; and 94.8%, 94.4%-95.2%, respectively). All of these estimates were significantly different (all P < .01), except for the comparison between NHB and NHW persons (P = .83).\\nThose diagnosed with distant stage CRC experienced the biggest improvements in 1-year cause-specific survival probabilities over time. An increasing trend in 1-year survival probabilities is apparent for persons of all race and ethnic groups. For AIAN and NHB persons, the two groups with lowest initial 1-year survival probabilities, improvements continued at a consistent rate over time. For API, Hispanic, and NHW persons, there appears to be a slowing of the rate of improvement in the second half of the study period. The adjusted time-averaged survival probabilities were low: 46.6% (42.3%-50.9%) for NHB, 48.3% (37.9%-58.7%) for AIAN, 50.5% (48.5%-52.5%) for NHW, 58.1% (54.2%-62.0%) for Hispanic, and 58.3% (54.2%-62.4%) for API persons. All pairwise comparisons among these groups were statistically significant (all P < .01), except for differences between the two groups with the lowest (AIAN vs NHB, P = .87), and between the two groups with the highest (API vs Hispanic, P = .69), average adjusted 1-year cause-specific survival probabilities.\",\n \"Estimates of adjusted 5-year cause-specific survival probabilities are shown for all race/ethnicity by stage combinations in Tables 4 and 5. Neither the race/ethnicity by stage by year of diagnosis three-way interaction (P = .90) nor the two-way interaction between the year of diagnosis and race/ethnicity (P = .59) were statistically significant; there is not sufficient evidence to conclude that there are race- or ethnic-specific differences in 5-year survival trends either within or across stages of CRC at diagnosis. There were statistically significant interactions between race/ethnicity and stage (P < .001), and between stage and the year of diagnosis trends (P < .001) with respect to differences in 5-year survival (see Figure 3).Table 4.Unadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199283.2 (67.0-99.3)58.1 (40.1-76.0)8.5 (0-29.1)49.9 (39.3-60.5)AIAN199587.5 (70.6-100)64.7 (47.7-81.7)13.4 (0-33.0)55.2 (44.9-65.5)AIAN199885.4 (69.9-100)73.1 (57.2-89.0)1.6 (0-18.5)53.3 (44.0-62.6)AIAN200291.8 (79.0-100)70.4 (56.5-84.3)10.8 (0-33.9)57.6 (47.7-67.6)AIAN200597.1 (85.6-100)65.6 (52.5-78.6)23.0 (6.8-39.2)61.9 (54.0-69.8)AIAN200897.5 (84.4-100)83.4 (70.7-96.1)5.9 (0-22.7)62.3 (54.0-70.6)AIAN201189.3 (77.0-100)67.5 (55.3-79.6)9.5 (0.0-22.7)55.4 (48.2-62.7)AIAN201496.5 (86.4-100)71.4 (60.1-82.6)15.1 (2.0-28.2)61.0 (54.3-67.6)API199297.1 (93.4-100)63.9 (57.9-70.0)7.1 (.9-13.2)56.0 (52.9-59.1)API199594.8 (91.1-98.4)67.5 (62.0-73.0)4.9 (0-10.7)55.7 (52.8-58.6)API199896.1 (93.0-99.2)76.8 (71.8-81.8)7.9 (2.5-13.4)60.3 (57.6-62.9)API200296.7 (94.3-99.1)79.8 (75.4-84.1)7.1 (2.5-11.8)61.2 (58.9-63.5)API200596.5 (94.1-98.9)80.0 (75.9-84.1)9.8 (4.7-14.8)62.1 (59.8-64.4)API200897.7 (95.7-99.7)80.6 (76.3-85)9.0 (4.1-13.9)62.4 (60.1-64.7)API201196.7 (94.4-99.0)76.7 (72.5-81.0)8.4 (3.6-13.1)60.6 (58.3-62.9)API201497.2 (95.0-99.4)81.5 (77.6-85.5)8.0 (3.8-12.1)62.2 (60.2-64.3)Hispanic199292.6 (87.7-97.5)58.1 (51.7-64.4)6.7 (0-13.4)52.4 (48.9-55.9)Hispanic199590.7 (85.5-95.8)61.0 (54.6-67.4)5.4 (0-11.4)52.4 (49.0-55.8)Hispanic199894.3 (90.6-98.0)72.7 (67.0-78.3)5.5 (0-11.5)57.5 (54.4-60.5)Hispanic200293.8 (90.1-97.4)75.5 (70.3-80.8)6.0 (.8-11.2)58.4 (55.7-61.2)Hispanic200594.6 (91.6-97.6)74.1 (69.3-78.9)9.3 (4.1-14.5)59.3 (56.8-61.9)Hispanic200896.0 (93.5-98.6)74.0 (69.1-78.9)6.3 (1.6-10.9)58.8 (56.4-61.2)Hispanic201197.6 (95.5-99.6)79 (74.6-83.3)8.9 (4.6-13.1)61.8 (59.6-63.9)Hispanic201496.9 (94.8-99)75.2 (70.9-79.4)9.4 (5.1-13.7)60.5 (58.4-62.6)NHB199290.6 (85.7-95.4)60.9 (54.1-67.7)2.2 (0-6.8)51.2 (48.1-54.4)NHB199591.3 (86.7-95.9)63.8 (57.0-70.6)4.3 (0-9.6)53.1 (49.8-56.4)NHB199896.0 (92.5-99.4)66.4 (60.0-72.9)4.8 (0-9.8)55.7 (52.8-58.7)NHB200289.9 (85.2-94.6)71.6 (65.9-77.3)4.5 (0-9.0)55.3 (52.4-58.2)NHB200591.8 (87.9-95.7)73.5 (67.7-79.3)4.3 (.1-8.4)56.5 (53.8-59.2)NHB200893.8 (90.4-97.1)69.3 (63.3-75.3)5.7 (1.4-10.1)56.3 (53.6-59.0)NHB201193.7 (90.2-97.2)68.7 (62.4-75.0)5.4 (1.2-9.5)55.9 (53.2-58.7)NHB201494.7 (91.4-98.0)69.2 (63.1-75.4)5.9 (1.7-10.1)56.6 (53.9-59.3)NHW199292.4 (91.1-93.7)64.4 (62.2-66.5)4.8 (3.2-6.5)53.9 (52.9-54.9)NHW199591.8 (90.5-93.2)65.6 (63.3-67.8)5.9 (4.2-7.6)54.4 (53.4-55.5)NHW199893.2 (92-94.4)67.8 (65.7-69.9)4.2 (2.7-5.7)55.1 (54.1-56)NHW200293.7 (92.6-94.9)73.4 (71.3-75.5)4.1 (2.7-5.5)57.1 (56.2-58)NHW200595.3 (94.2-96.3)77.7 (75.7-79.7)5.1 (3.5-6.7)59.4 (58.4-60.3)NHW200895.4 (94.4-96.4)77.6 (75.6-79.7)9.0 (7.0-11.1)60.7 (59.7-61.7)NHW201196.1 (95.1-97.2)78.8 (76.7-80.9)4.8 (3.3-6.4)59.9 (59.0-60.9)NHW201496.2 (95.1-97.3)79.3 (77.2-81.4)8.1 (6.2-10.1)61.2 (60.2-62.2)Table 5.Adjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.LocalizedRegionalDistantOverallRace/ethnicityYear of DxSurvival (95% CI)Survival (95% CI)Survival (95% CI)Survival (95% CI)AIAN199278.8 (64.3-93.4)56.6 (40.4-72.7)7.5 (.0-26.0)47.6 (38.1-57.1)AIAN199585.5 (70.3-100)62.1 (46.8-77.4)13.3 (.0-31.0)53.7 (44.4-63.0)AIAN199881.1 (67.1-95.0)70.0 (55.6-84.3)2.8 (.0-18.0)51.3 (42.8-59.7)AIAN200289.4 (77.9-100)68.8 (56.3-81.3)8.2 (.0-29.1)55.5 (46.5-64.5)AIAN200595.6 (85.2-100)63.8 (52.0-75.6)20.4 (5.8-35.0)59.9 (52.7-67.1)AIAN200893.6 (81.8-100)79.8 (68.3-91.2)4.3 (.0-19.5)59.2 (51.8-66.7)AIAN201184.9 (73.8-96.0)65.4 (54.4-76.3)7.3 (.0-19.2)52.5 (46.0-59.1)AIAN201493.0 (84.0-100)68.6 (58.4-78.7)13.3 (1.5-25.1)58.3 (52.3-64.3)API199295.4 (92.1-98.8)63.1 (57.6-68.5)6.5 (.9-12.0)55.0 (52.1-57.9)API199592.3 (88.9-95.7)67.1 (62.1-72.1)6.4 (1.1-11.6)55.3 (52.5-58.0)API199895.2 (92.4-98.1)75.7 (71.2-80.2)8.8 (3.8-13.7)59.9 (57.4-62.4)API200294.5 (92.2-96.7)78.1 (74.2-82.1)7.9 (3.7-12.1)60.2 (58.0-62.3)API200593.7 (91.4-96.0)78.3 (74.6-82.1)10.6 (6.0-15.2)60.9 (58.6-63.1)API200895.3 (93.3-97.2)78.9 (74.9-82.9)10.7 (6.2-15.2)61.6 (59.4-63.8)API201193.2 (90.9-95.4)75.4 (71.5-79.3)9.3 (5.0-13.7)59.3 (57.1-61.5)API201492.6 (90.5-94.7)78.9 (75.2-82.5)11.6 (7.8-15.4)61.0 (59.0-63.0)Hispanic199289.7 (85.3-94.2)57.3 (51.5-63.0)8.6 (2.5-14.8)51.9 (48.6-55.1)Hispanic199588.8 (84.1-93.5)60.2 (54.4-66.0)5.6 (.2-11.0)51.5 (48.4-54.7)Hispanic199891.2 (87.8-94.6)71.2 (66.1-76.4)6.5 (1.1-12.0)56.3 (53.5-59.1)Hispanic200291.0 (87.7-94.4)73.6 (68.8-78.4)5.8 (1.0-10.5)56.8 (54.2-59.4)Hispanic200591.5 (88.7-94.4)72.7 (68.3-77.0)8.8 (4.0-13.5)57.7 (55.2-60.1)Hispanic200892.6 (90.2-95.0)71.6 (67.2-76.1)6.6 (2.3-10.8)57.0 (54.7-59.2)Hispanic201194.6 (92.6-96.6)76.7 (72.7-80.7)8.4 (4.5-12.3)59.9 (57.8-62.0)Hispanic201492.5 (90.4-94.5)72.8 (68.9-76.7)9.6 (5.7-13.6)58.3 (56.2-60.4)NHB199287.5 (83.0-91.9)59.6 (53.4-65.8)1.3 (.0-5.5)49.4 (46.5-52.4)NHB199589.5 (85.2-93.7)62.3 (56.2-68.5)3.4 (.0-8.3)51.7 (48.7-54.8)NHB199892.7 (89.5-95.9)64.6 (58.7-70.4)5.4 (.9-10.0)54.2 (51.5-57.0)NHB200286.8 (82.5-91.0)69.7 (64.5-74.9)4.2 (.0-8.4)53.6 (50.8-56.3)NHB200589.0 (85.4-92.6)71.5 (66.2-76.8)3.7 (.0-7.5)54.7 (52.2-57.3)NHB200890.0 (86.9-93.2)67.7 (62.2-73.1)6.2 (2.1-10.2)54.6 (52.1-57.2)NHB201190.0 (86.7-93.2)66.8 (61.1-72.6)4.0 (.1-7.8)53.6 (51.0-56.2)NHB201490.9 (87.8-94.0)68.0 (62.4-73.6)6.1 (2.2-10.0)55.0 (52.4-57.6)NHW199290.7 (89.3-92.1)65.1 (63.0-67.2)7.3 (5.7-9.0)54.4 (53.2-55.6)NHW199590.5 (89.1-92.0)66.4 (64.3-68.6)8.0 (6.3-9.8)55.0 (53.8-56.2)NHW199891.1 (89.8-92.4)68.6 (66.6-70.7)9.4 (7.8-10.9)56.4 (55.2-57.5)NHW200292.1 (90.8-93.4)73.5 (71.5-75.6)8.7 (7.2-10.2)58.1 (57.0-59.3)NHW200592.3 (91.0-93.5)77.1 (75.1-79.0)10.7 (9.0-12.3)60.0 (58.9-61.1)NHW200892.7 (91.5-93.9)77.1 (75.0-79.1)12.3 (10.3-14.3)60.7 (59.5-61.9)NHW201193.5 (92.2-94.7)78.0 (75.9-80.0)10.7 (9.1-12.3)60.7 (59.6-61.9)NHW201492.3 (91.1-93.6)77.7 (75.7-79.8)13.2 (11.3-15.1)61.1 (59.9-62.3)Figure 3.Trends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nUnadjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nAdjusted 5-Year Cause-Specific Survival Probabilities (%), by Race and Ethnicity, Stage, and Year of Diagnosis.\\nTrends in 5-year adjusted survival probabilities for individuals from racial and ethnic subgroups diagnosed at different stages of colorectal cancer from 1992 through 2014.\\nThe trends reflect improvements in 5-year cause-specific survival probabilities over time for all CRC stages. The rate of improvement in 5-year survival probabilities over time was greatest for those with regional stage CRC, although the rate of improvement appears to be slowing in the second half of the study period. As there are no significantly different trends over time in adjusted 5-year survival probabilities by race or ethnicity, key differences among race and ethnic groups are best summarized by differences in their adjusted time-averaged survival probabilities. For those diagnosed at localized stage CRC, estimates of average adjusted 5-year survival probabilities were 86.9% (84.2%-89.6%) for AIAN, 90.2% (89.4%-91.0%) for NHB, 92.0% (91.8%-92.2%) for NHW, 92.1% (91.5%-92.7%) for Hispanic, and 94.1% (93.5%-94.7%) for API. These differed significantly between all pairs of groups (all P < .04), except for the Hispanic vs NHW comparison (P = .56). For those diagnosed at regional stage CRC, estimates of average adjusted 5-year survival probabilities were 65.1% (62.2%-68.0%) for NHB, 67.5% (61.0%-74.0%) for AIAN, 70.5% (68.1%-72.9%) for Hispanic, 72.5% (71.5%-73.5%) for NHW, and 75.1% (72.9%-77.3%) for API persons. These differed significantly between all pairs of groups (all P < .05), except for the AIAN vs NHB comparison (P = .22). For those diagnosed at distant stage CRC, estimates of average adjusted 5-year survival probabilities were 4.6% (3.6%-5.6%) for NHB, 7.8% (6.6%-9.0%) for Hispanic, 8.5% (7.3%-9.7%) for API, 9.0% (5.5%-12.5%) for AIAN, and 9.6% (9.2%-10.0%) for NHW. The AIAN average did not differ significantly from those of the API, Hispanic, or NHW groups, nor did the API vs Hispanic averages. All others differed significantly (all P < .01).\",\n \"The purpose of this study was to estimate and compare race- and ethnicity-specific 1- and 5-year CRC cause-specific survival trends, within the context of the stage at which the cancer was first detected. We utilized data from the SEER registries for persons diagnosed with CRC from 1992 to 2018 and estimated 1- and 5-year survival probabilities within groups of persons with CRC by race and ethnicity, stage of disease, and year of diagnosis. Study findings contribute to existing knowledge gaps regarding trends in survival by including estimates for an often-overlooked population: American Indian or Alaska Natives.\\nThere are significant differences in the degree to which improvements in 1-year survival probabilities following stage-specific CRC are experienced according to a person’s race and ethnicity. The estimates that we report specifically for AIAN and API subgroups contribute towards the limited literature focused on identifying survival trend differences for these racially minoritized populations. Study findings suggest that AIAN individuals have the lowest 1-year survival probabilities compared to those of other races. Lower 1-year CRC-specific survival probabilities are also seen among NHB persons. In particular, these persons have the lowest average survival probabilities following the diagnosis of distant-stage CRC. Although differences among racial and ethnic groups have moderated somewhat over the study period, these disparities do not appear to have resolved completely.\\nThere is established evidence that CRC screening can prevent or detect CRC early.27,28 For screening to be effective in improving outcomes, timely follow up of any abnormal test is necessary. Structural barriers (eg, lack of insurance or social support, racism and discrimination) to obtaining appropriate primary care services may play an important role in reducing the possibility for racially minoritized populations to receive guideline-compliant screening services.5,29-31 The fact that some studies conducted within the Veterans’ Health Administration found no differences between White and Black persons in diagnostic follow up testing,32,33 suggests that access to appropriate structures and services may play an important role in appropriate post-screening follow-up for minoritized populations. As expected, the greatest differences in survival probabilities are apparent among individuals diagnosed at different stages of CRC. Although these patterns are largely similar among persons from different race and ethnic groups disparities in survival persist. Enhanced follow-up of abnormal results, may help overcome at least some of the persistent disparities in CRC survival probabilities among persons from different race and ethnic groups.\\nOther differences in access to and utilization of quality health care may also contribute to the observed differences of survival trends by race.34-36 Some studies have suggested NHB persons are less likely to receive surgical treatment and adjuvant chemotherapy.35-38 Social, structural and political determinants also contribute towards the likelihood of NHB and AIAN persons to be diagnosed with advanced stage CRC compared to NHW persons.39,40 This may further contribute to the lower survival probabilities observed in NHB persons with distant stage CRC.41,42 Racial and ethnic minoritized populations also experience disparities in terms of post-treatment surveillance and distinct baseline comorbidities, which further contribute to lower survival rates.34,35 Several studies continue to highlight that transportation barriers, cultural beliefs, fear and stigma about screening, and concerns about privacy issues are contributing factors to survival outcomes43-47 for persons with CRC.\\nLifestyle and biological factors may also play a role in CRC risk and outcomes. For instance, lifestyle factors such as obesity, smoking, alcohol consumption, and physical inactivity have a higher prevalence among Black populations.48 Also, genetic mutations and microsatellite instability can differ among racially and ethnically minoritized populations. All of these things can affect CRC development and prognosis,49-54 and may play into the disparities noted in this work.\\nThere are several potential limitations in our analysis. First, we must acknowledge that this is a retrospective analysis of data from the SEER registries. However, SEER is a population-based resource which provides information about the most critical factors of interest, and we tried to control for these. Second, we cannot fully exclude the possibility that regional variations may contribute to the observed CRC survival probability estimates that average across SEER registries. Third, we chose to report on cause-specific survival, and this may be influenced by different practices in cause of death ascertainment. However, in sensitivity analyses performed using relative survival estimates we found that overall findings were largely concordant. This is similar to a report of data from Canada where differences in survival probabilities were noted, and that differences between First Nations and non-aboriginal persons were somewhat smaller for cause-specific survival when compared to relative survival estimates.55 Fourth, SEER data have relatively little information on comorbidities, access to care, and insurance status. Fifth, racial and ethnic classifications in medical records may reflect misclassification,56 which may bias our estimates. Finally, SEER registries represent a subset of AIAN persons spanning the United States, but do not capture data from some regions with large AIAN populations, including those in Oklahoma, Arizona, or the Northern Plains and Great Lakes.57 Even with these limitations, we are able to provide new data concerning trends in CRC survival over time for multiple racial and ethnic groups in the United States, according to their stage at diagnosis.\\nIn conclusion, we present trends in 1- and 5-year CRC cause-specific survival probabilities for persons of five racial and ethnic groups. These estimates of, and trends in, survival probabilities for groups of minoritized races and ethnicities may enable the development of a more complete picture of CRC prognosis. We have identified significant differences in the race- and ethnic-specific trends in 1-year stage-specific CRC survival probabilities. In particular, AIAN persons have historically experienced poorer CRC prognosis, as have NHB individuals. Although these disparities appear to be lessening somewhat, the current differences in survival probabilities continue to call for further work in order to erase them completely. Trends in 1-year survival probabilities are significantly different among those diagnosed at different stages of CRC. This, coupled with the finding that the distribution of stages at diagnosis appears to differ among persons of different racial and ethnic groups, suggests that prioritizing specific stages of disease at diagnosis may provide 1 avenue that may help us to overcome race- and ethnic-specific disparities in CRC-specific survival probability. This also argues for a potential need to improve screening for CRC, as this may influence the distribution of stages at which it is diagnosed. Future research should strive to capture cancer incidence and survival information from all key racial and ethnic subgroups. Future efforts should also incorporate and evaluate multi-level interventions at the individual, structural, and policy levels to address the persistent disparities in CRC survival.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["intro","methods",null,null,null,"results",null,null,null,"discussion"],"string":"[\n \"intro\",\n \"methods\",\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\"\n]"},"keywords":{"kind":"list like","value":["colorectal cancer","cause-specific survival","trends","minority health","disparities"],"string":"[\n \"colorectal cancer\",\n \"cause-specific survival\",\n \"trends\",\n \"minority health\",\n \"disparities\"\n]"}}},{"rowIdx":364,"cells":{"title":{"kind":"string","value":"Spatial Analysis of Breast Cancer Mortality Rates in a Rural State."},"pmid":{"kind":"string","value":"36265079"},"background_abstract":{"kind":"string","value":"Breast cancer affects 1 in 8 women in the US and is the most frequently diagnosed cancer in women. In South Dakota, 102 women die from breast cancer each year. We assessed which sociodemographic factors contributed to mortality rates in South Dakota and used spatial analysis to investigate how counties' observed age-adjusted mortality rates compared with expected rates."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"We computed standardized incidence ratios (SIRs) of all counties in South Dakota by using the age-adjusted mortality rates, the 2000 US standard population, and the South Dakota estimated population. We used a linear regression model to identify sociodemographic factors associated with breast cancer mortality rates and to compute a new SIR value, after controlling for relevant factors."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Educational level and breast cancer incidence rates were significantly associated with breast cancer mortality rates at the county level. The SIR values based on age-adjusted counts showed which counties had more deaths due to breast cancer than what might be expected using South Dakota as the reference population. After controlling for sociodemographic factors, the range of SIR values decreased and had lower variability."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The regression model helped identify factors associated with mortality and provided insights into which risk factors are at play in South Dakota. This information, in combination with the spatial distribution of mortality by county, can be used to help allocate resources to the counties in South Dakota that need them most."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Female","Breast Neoplasms","Incidence","Rural Population","Risk Factors","Spatial Analysis"],"string":"[\n \"Humans\",\n \"Female\",\n \"Breast Neoplasms\",\n \"Incidence\",\n \"Rural Population\",\n \"Risk Factors\",\n \"Spatial Analysis\"\n]"},"pmcid":{"kind":"string","value":"9616131"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"In South Dakota, breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among women (1,2). In 2022, an estimated 750 new cases and 110 deaths attributed to female breast cancer will occur in South Dakota. In general, a woman in the US has a 1-in-8 lifetime risk of developing breast cancer (3,4). Since 1989, the US breast cancer mortality rate has decreased 40%, but from 2010 to 2019 the rate slowed to a low of decreasing by 1.3% per year (5).\nCharacteristics such as age and race and ethnicity affect a woman’s chances of being diagnosed with or dying of breast cancer, but new evidence has established that sociodemographic factors, including education level, also play a role (6). Albano et al noted a negative relationship between number of years of education and cancer mortality and found that the level of education and race vary considerably with mortality rates (7). Of the South Dakota population aged 25 years or older, 92.2% are high school graduates (higher than the national average) and 29.3% have a bachelor’s degree or higher (lower than national average) (8). Olson et al acknowledged that communities exist in which geographic disparities are more prominent because of rural isolation and small population size (9). Furthermore, 64 of 66 counties in South Dakota are categorized as rural or frontier, and South Dakota contains 9 American Indian reservations (10). Finally, 61.6% of women receiving breast services are White, and 16.7% are American Indian (11); most of the population in South Dakota is White, and the leading minority is American Indian at 8.8% (8).\nThe study aimed to describe the spatial distribution of female breast cancer mortality at the county level in South Dakota and assess the association between mortality rates and risk factors reported in the literature."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Data source [SUBSECTION] The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\nThe 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\n[SUBTITLE] Cancer data [SUBSECTION] Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\nBreast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\n[SUBTITLE] Demographics [SUBSECTION] We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\nWe used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\n[SUBTITLE] Statistical analysis [SUBSECTION] [SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\n[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\nA SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24)."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Exploratory data analysis [SUBSECTION] To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\nTo learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\n[SUBTITLE] Regression analysis [SUBSECTION] Model 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.\nModel 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] Thirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values.\nThirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Data source","Cancer data","Demographics","Statistical analysis","Data manipulation and missing value imputation","Multiple linear regression","Standardized incidence ratio","Exploratory data analysis","Regression analysis","Standardized incidence ratio"],"string":"[\n \"Data source\",\n \"Cancer data\",\n \"Demographics\",\n \"Statistical analysis\",\n \"Data manipulation and missing value imputation\",\n \"Multiple linear regression\",\n \"Standardized incidence ratio\",\n \"Exploratory data analysis\",\n \"Regression analysis\",\n \"Standardized incidence ratio\"\n]"},"other_sections_texts":{"kind":"list like","value":["The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%","Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).","We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).","[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).","Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).","We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).","A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).","To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.","Model 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.","Thirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values."],"string":"[\n \"The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\",\n \"Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\",\n \"We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \\nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\",\n \"[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\",\n \"Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\",\n \"We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\",\n \"A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\\n is the observed value for county j, and Ej\\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\\n is the jth\\n county’s observed age-adjusted mortality rate and yj\\n is the jth\\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\\n. All other analysis of the data was completed with the original variable (Oj\\n), as described in the data source.\\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\\n). The second SIR (SIRLM\\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\",\n \"To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\",\n \"Model 2 had the highest adjusted R\\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \\nAbbreviation: — , not applicable; AIC, Akaike information criterion.\\n\\nP = .01.\\n\\nP = .001.\",\n \"Thirty-five of the 66 counties had a SIRCOUNT\\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\\n was 2.15 and the lowest was 0.31.\\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\\nOn the other hand, for the SIRLM\\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\\n greater than 1. The 5 counties with the highest SIRLM\\n were similar to the SIRCOUNT\\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\\n was 1.94 and the lowest SIRLM\\n was 0.35. \\nThe results of the SIRs (SIRCOUNT\\n and SIRLM\\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\\n values. These counties with the highest SIRCOUNT\\n made up 5 of the 6 counties with the highest SIRLM\\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\\n values were the same 2 counties with the lowest SIRLM\\n values.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["What is known on this topic?","What is added by this report?","What are the implications for public health practice?","Introduction","Methods","Data source","Cancer data","Demographics","Statistical analysis","Data manipulation and missing value imputation","Multiple linear regression","Standardized incidence ratio","Results","Exploratory data analysis","Regression analysis","Standardized incidence ratio","Discussion"],"string":"[\n \"What is known on this topic?\",\n \"What is added by this report?\",\n \"What are the implications for public health practice?\",\n \"Introduction\",\n \"Methods\",\n \"Data source\",\n \"Cancer data\",\n \"Demographics\",\n \"Statistical analysis\",\n \"Data manipulation and missing value imputation\",\n \"Multiple linear regression\",\n \"Standardized incidence ratio\",\n \"Results\",\n \"Exploratory data analysis\",\n \"Regression analysis\",\n \"Standardized incidence ratio\",\n \"Discussion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Breast cancer affects 1 in 8 women in the US and is the most frequently diagnosed cancer in women.","Because South Dakota is a rural state, sociodemographic factors affect the population differently than in the general US population. We assessed the spatial distribution of breast cancer mortality rates by county, and our findings add insight on educational attainment as a risk factor for breast cancer.","Our results can be used to help allocate resources to the South Dakota counties that need them most.","In South Dakota, breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among women (1,2). In 2022, an estimated 750 new cases and 110 deaths attributed to female breast cancer will occur in South Dakota. In general, a woman in the US has a 1-in-8 lifetime risk of developing breast cancer (3,4). Since 1989, the US breast cancer mortality rate has decreased 40%, but from 2010 to 2019 the rate slowed to a low of decreasing by 1.3% per year (5).\nCharacteristics such as age and race and ethnicity affect a woman’s chances of being diagnosed with or dying of breast cancer, but new evidence has established that sociodemographic factors, including education level, also play a role (6). Albano et al noted a negative relationship between number of years of education and cancer mortality and found that the level of education and race vary considerably with mortality rates (7). Of the South Dakota population aged 25 years or older, 92.2% are high school graduates (higher than the national average) and 29.3% have a bachelor’s degree or higher (lower than national average) (8). Olson et al acknowledged that communities exist in which geographic disparities are more prominent because of rural isolation and small population size (9). Furthermore, 64 of 66 counties in South Dakota are categorized as rural or frontier, and South Dakota contains 9 American Indian reservations (10). Finally, 61.6% of women receiving breast services are White, and 16.7% are American Indian (11); most of the population in South Dakota is White, and the leading minority is American Indian at 8.8% (8).\nThe study aimed to describe the spatial distribution of female breast cancer mortality at the county level in South Dakota and assess the association between mortality rates and risk factors reported in the literature.","[SUBTITLE] Data source [SUBSECTION] The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\nThe 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\n[SUBTITLE] Cancer data [SUBSECTION] Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\nBreast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\n[SUBTITLE] Demographics [SUBSECTION] We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\nWe used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\n[SUBTITLE] Statistical analysis [SUBSECTION] [SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\n[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\nA SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).","The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%","Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).","We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).","[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).","Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\n with its k nearest neighbors and then approximates xi\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).","We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\n county and β^0 is the intercept. The variables X\n1\n\ni \nthrough Xji\n represent the values of the factors for the ith\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\n).","A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\n is the observed value for county j, and Ej\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\n is the jth\n county’s observed age-adjusted mortality rate and yj\n is the jth\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\n. All other analysis of the data was completed with the original variable (Oj\n), as described in the data source.\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\n). The second SIR (SIRLM\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).","[SUBTITLE] Exploratory data analysis [SUBSECTION] To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\nTo learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\n[SUBTITLE] Regression analysis [SUBSECTION] Model 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.\nModel 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] Thirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values.\nThirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values.","To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.","Model 2 had the highest adjusted R\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \nAbbreviation: — , not applicable; AIC, Akaike information criterion.\n\nP = .01.\n\nP = .001.","Thirty-five of the 66 counties had a SIRCOUNT\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\n was 2.15 and the lowest was 0.31.\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\nOn the other hand, for the SIRLM\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\n greater than 1. The 5 counties with the highest SIRLM\n were similar to the SIRCOUNT\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\n was 1.94 and the lowest SIRLM\n was 0.35. \nThe results of the SIRs (SIRCOUNT\n and SIRLM\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\n values. These counties with the highest SIRCOUNT\n made up 5 of the 6 counties with the highest SIRLM\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\n values were the same 2 counties with the lowest SIRLM\n values.","Overall, we found a significant association between incidence rate and educational level with respect to breast cancer mortality rates. Breast cancer incidence was positively associated with mortality rates in South Dakota, which suggests that more breast cancer cases are associated with more breast cancer deaths. In addition, educational attainment was repeatedly identified as a significant factor for mortality. Gadeyne et al found inconclusive results in their study of breast cancer mortality and education; however, Albano et al found a significant association between educational levels and cancer in general (7,25), specifically that lower educational attainment was related to higher cancer mortality rates, reflecting the findings of this study. Race, median age, and number of women screened were not selected in the feature selection during stepwise regression in our study; similarly, race, median age, and number of women screened were not significant in our full model.\nThe 2020–2021 South Dakota Department of Education yearly review stated that American Indians were the largest minority group in school. However, American Indians still have a 63% completion rate for high school graduation and 59% attendance rate, compared with Whites who have a 94% completion rate for high school graduation and 94% attendance rate (26). An interesting point to consider is that South Dakota has no set standards for sex education (27). Thus, students are not taught reproductive health in general, including the importance of breast examinations, Pap smears, or prostate examinations. We advocate that set and scientifically backed health standards in high school would expose students at an early age to the risks of breast cancer and their options for screening.\nThe western half of South Dakota had more variability in SIR values, and the state’s demographics could be a possible explanation. The 4 counties with the highest SIRs for both count and linear model SIRs were Corson, Perkins, Aurora, and Mellette, which are either in an American Indian reservation or neighbor a county within an American Indian reservation. Research on 3 tribes in western South Dakota supported that trust is often a barrier for American Indians (as are remote location and approvals by Indian Health Service programs) (28). Research in New Mexico reported that even after in-depth implementation of screening programs that lowered the barriers of cost, availability, and access to Native American and Hispanic women, the screening rates remained low, under 40% of women annually (26). The high SIRs in or neighboring reservation counties may mean that trust is also an issue, and South Dakota has more to work on than accessibility to Native Americans.\nAfter controlling for incidence rate and educational attainment, the SIRLM\n values became less variable. The SIRs’ decrease in range and mean closer to 1 indicate that the factors did affect mortality rate. This again agrees with findings from Albano et al that educational attainment affects mortality (7).\nWe found that some counties had a higher mortality rate than expected based on the age of the women in the county. Ziebach and Jackson counties had the highest mortality rates, and the counties with the lowest SIRs are not home to major medical centers. Haakon County is vertically between Ziebach and Jackson counties and is one of the counties that does not have a provider; however, Haakon County has a higher mortality rate than the average of counties of South Dakota and both SIRs greater than 1, which means there were more deaths than expected. The areas of the map where there are dark green counties next to dark red counties are either on an American Indian reservation or neighbor an American Indian reservation. The differences between counties do not come from any singular cause, but rather due to variations in race, poverty levels, and population size.\nOur study has limitations, primarily in the absence of portions of mortality data. Because South Dakota is largely a rural state, several counties have small populations and see very few deaths from breast cancer. These numbers are then withheld from the public to protect the privacy of the patients. This suppression resulted in having to impute the mortality rates of 15 counties, possibly introducing errors. The assumptions of the regression model did not account for this error, which may confer bias on the results. The counties with American Indian reservations have another health system that could have resulted in the under-representation or over-representation of breast cancer deaths in those counties (29). In addition, a study found that the misclassification of the race of Native Americans caused an underestimation of mortality rates as well (30). These gaps in databases and their contents highlight research challenges that rural communities will continue to face when few data are collected, populations and incidence are sparse, and data are inconsistently collected by multiple sources. Sociodemographic data are also challenging to consistently collect throughout a state. More detailed data per county would help yield accurate and unbiased results. For example, considering education, Zajacova and Lawrence argue that education is not a single-generation factor (31). Having data on the educational attainment of a patient’s parents or family, in addition to their own educational attainment, would allow us to assess the risk and see the relationship between education and breast cancer incidence or mortality. Thus, more research is needed to understand the effects education level has on financial security, stable employment, social success, and in turn, breast cancer mortality.\nIn conclusion, understanding the risk factors and geographic distribution of breast cancer mortality among women across the state will assist stakeholders with efforts at prevention and resource allocation guided by data."],"string":"[\n \"Breast cancer affects 1 in 8 women in the US and is the most frequently diagnosed cancer in women.\",\n \"Because South Dakota is a rural state, sociodemographic factors affect the population differently than in the general US population. We assessed the spatial distribution of breast cancer mortality rates by county, and our findings add insight on educational attainment as a risk factor for breast cancer.\",\n \"Our results can be used to help allocate resources to the South Dakota counties that need them most.\",\n \"In South Dakota, breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among women (1,2). In 2022, an estimated 750 new cases and 110 deaths attributed to female breast cancer will occur in South Dakota. In general, a woman in the US has a 1-in-8 lifetime risk of developing breast cancer (3,4). Since 1989, the US breast cancer mortality rate has decreased 40%, but from 2010 to 2019 the rate slowed to a low of decreasing by 1.3% per year (5).\\nCharacteristics such as age and race and ethnicity affect a woman’s chances of being diagnosed with or dying of breast cancer, but new evidence has established that sociodemographic factors, including education level, also play a role (6). Albano et al noted a negative relationship between number of years of education and cancer mortality and found that the level of education and race vary considerably with mortality rates (7). Of the South Dakota population aged 25 years or older, 92.2% are high school graduates (higher than the national average) and 29.3% have a bachelor’s degree or higher (lower than national average) (8). Olson et al acknowledged that communities exist in which geographic disparities are more prominent because of rural isolation and small population size (9). Furthermore, 64 of 66 counties in South Dakota are categorized as rural or frontier, and South Dakota contains 9 American Indian reservations (10). Finally, 61.6% of women receiving breast services are White, and 16.7% are American Indian (11); most of the population in South Dakota is White, and the leading minority is American Indian at 8.8% (8).\\nThe study aimed to describe the spatial distribution of female breast cancer mortality at the county level in South Dakota and assess the association between mortality rates and risk factors reported in the literature.\",\n \"[SUBTITLE] Data source [SUBSECTION] The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\\nThe 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\\n[SUBTITLE] Cancer data [SUBSECTION] Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\\nBreast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\\n[SUBTITLE] Demographics [SUBSECTION] We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \\nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\\nWe used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \\nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\\n[SUBTITLE] Statistical analysis [SUBSECTION] [SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\n[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\\n is the observed value for county j, and Ej\\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\\n is the jth\\n county’s observed age-adjusted mortality rate and yj\\n is the jth\\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\\n. All other analysis of the data was completed with the original variable (Oj\\n), as described in the data source.\\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\\n). The second SIR (SIRLM\\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\\nA SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\\n is the observed value for county j, and Ej\\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\\n is the jth\\n county’s observed age-adjusted mortality rate and yj\\n is the jth\\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\\n. All other analysis of the data was completed with the original variable (Oj\\n), as described in the data source.\\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\\n). The second SIR (SIRLM\\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\",\n \"The 66 counties of South Dakota have boundaries that are defined by the South Dakota Legislature and accepted by the US Census. The counties range in population from 183,439 in Minnehaha County to 917 in Jones County, and the median population per county is 5,413. Most residents of South Dakota were White; the median percentage of non-White residents by county was 6.6% and the maximum was 95.2%\",\n \"Breast cancer incidence and mortality rates from 2001 through 2015 were extracted from the South Dakota State Cancer Registry provided by the South Dakota Department of Health (12). Both rates were per 100,000 people and age-adjusted to the 2000 US standard population and the South Dakota estimated population. The proportion of mammography screening rates in South Dakota was based on the numbers reported by Holzhauser et al for the All Women Count! mammography program (13). The average number of participants for 1997 through 2016 was reported for the program; then, the average number of participants was adjusted for the total number of women older than 40 years in the county to get an estimated screening rate for each county (13,14).\",\n \"We used 2015 data from the US Census Bureau to obtain information on the 66 South Dakota counties, including the number of providers and the education level, poverty level, percentage of uninsured, median age, and race of residents (8).\\nData on educational attainment were obtained from the US Census Bureau’s American Community Survey (ACS). These data were count estimates for the population of each county aged 25 years or older. Levels were categorized as less than 9th grade, 9th through 12th grade with no diploma, high school graduate or equivalent, some college but no degree, associate degree, bachelor’s degree, and graduate or professional degree. These values were modified into an educational attainment statistic of the percentage of the population with less than a bachelor’s degree of education. The statistic used in this study was the percentage of the population with less than a bachelor’s degree, by county.\\nWe collected data on poverty estimates, by county, from the ACS; these data adhered to the standards specified by the Office of Management and Budget in Statistical Policy Directive 14 (8). Poverty was determined by a set of income thresholds that consider the living situation (alone or with nonrelatives), age, and number of people per household. For example, the poverty threshold for 2-person families varies by the age of the primary householder and differs from the poverty threshold for people living alone or with nonrelatives, which also varies by age. \\nInsurance coverage percentages were collected from Small Area Health Insurance Estimates (SAHIE) (15). The uninsured percentage included residents who were not covered by insurance, which excluded those on government assistance such as Medicaid or Medicare. Finally, the data set summarizing racial distributions in a county included estimated population counts for American Indian and Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, other race, and 2 or more races. Because of South Dakota's predominantly White population, the data were configured into White and non-White, which determined the non-White percentage per county (8) (Table 1).\",\n \"[SUBTITLE] Data manipulation and missing value imputation [SUBSECTION] Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\nMortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\\n[SUBTITLE] Multiple linear regression [SUBSECTION] We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\\nWe used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\",\n \"Mortality rates and the various independent variables were combined into 1 data set; 15 of the 66 counties were missing mortality rate data. Mortality rates are often suppressed from public availability when 3 or fewer deaths are reported in a county, to protect patient identity. To remedy the missing data, k-nearest-neighbor (KNN) imputation was used to estimate the missing mortality values. KNN imputation compares a data point xi\\n with its k nearest neighbors and then approximates xi\\n using the majority vote of these k neighbors in multidimensional space. For the data, k = 9 nearest neighbors were used, and a weighted mean of the k nearest values was placed for each missing xi \\n(16,17). This was done with the function “knn()” from the R package VIM version 6.1.1 (R Foundation for Statistical Computing) (18).\",\n \"We used multiple linear regression to model the relationship between the factors in this study and breast cancer mortality rates (19). We considered several potential predictor variables with observed correlation, hence a stepwise variable selection technique was used, in both the forward and backward direction, to perform feature selection. As a result, a subset of the factors that were associated with the mortality rates was obtained based on Akaike information criterion (AIC) (19,20).\\nThe resulting model is of the form Yi ^= β^0 + β^1X1i+…+β^pXji, for i = 1,2, … ,66 where Yi ^ represents the estimated mortality rate for the ith\\n county and β^0 is the intercept. The variables X\\n1\\n\\ni \\nthrough Xji\\n represent the values of the factors for the ith\\n county and β^1 through β^p are the coefficients that were estimated using the least squares regression method (19,21). To explore the data, 4 linear regression models were created, which differ by factors included in the model.\\nThe first regression model included all 8 factors as prediction variables resulting in Model 1. The regression model was then fitted by stepwise variable selection in both the forward and backward direction by using AIC as a model selection criterion. AIC rewards goodness of fit and penalizes the model’s complexity (19). This was done by using the R package MASS version 7.3–54 (R Foundation for Statistical Computing) (22). The simplified model resulted in Model 2. At this point, a decision was made to remove the incidence rate from the data to better see how the other sociodemographic factors contributed to breast cancer outcomes, resulting in Model 3. Model 3 was then fitted with stepwise variable selection, yielding Model 4. To best compare the expected mortality to the observed mortality with all predictors available, Model 2 was chosen as the final model. Model 2 was then used to predict the expected mortality rates for the second standardized incidence ratio (SIR) that was computed (SIRLM\\n).\",\n \"A SIR was used to compare the spatial distribution of counties in terms of mortality rates due to breast cancer. In general, the SIR compares the expected value of deaths to the observed value of deaths in a county. This is calculated with SIRj=OjEj where Oj\\n is the observed value for county j, and Ej\\n is the expected value for county j. A SIR of greater than 1 means that there were more deaths than expected, whereas a SIR of less than 1 means that there were fewer deaths than expected for that county.\\nBecause age-specific mortality rates by county were not available, the observed age-adjusted counts per county were computed from the age-adjusted rates per county as follows: Oj*=Oj100,000yj and Ej*=∑j=166Oj100,000·yj·yj∑j=166yj , where Oj\\n is the jth\\n county’s observed age-adjusted mortality rate and yj\\n is the jth\\n county’s population. The expected count was computed to be on the same scale as the observed count. These manipulations were used only to calculate an age-adjusted count SIR, referred to as SIRCOUNT\\n. All other analysis of the data was completed with the original variable (Oj\\n), as described in the data source.\\nA SIR was calculated on the mortality in each South Dakota county (N = 66). To account for the age adjustment of the data, 2 different SIRs were found: using the age-adjusted mortality count for South Dakota and using the expected rate obtained from the linear regression model. The first SIR accounted only for the age adjustment, and the second SIR accounted for more factors related to breast cancer. For example, the first SIR used the age-adjusted mortality count for a county for both observed and expected (SIRCOUNT\\n). The second SIR (SIRLM\\n) used the mortality rates by county, which were per 100,000 persons and age-adjusted to the 2000 US standard population, and the South Dakota estimated population for observed and predicted mortality rate from the linear regression model for expected, which accounted for incidence rates and educational attainment. All statistical analyses used R version 4.1.2 and RStudio version 2021.09.2 build 382 (R Foundation for Statistical Computing) (23,24).\",\n \"[SUBTITLE] Exploratory data analysis [SUBSECTION] To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\\nTo learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\\n[SUBTITLE] Regression analysis [SUBSECTION] Model 2 had the highest adjusted R\\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \\nAbbreviation: — , not applicable; AIC, Akaike information criterion.\\n\\nP = .01.\\n\\nP = .001.\\nModel 2 had the highest adjusted R\\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \\nAbbreviation: — , not applicable; AIC, Akaike information criterion.\\n\\nP = .01.\\n\\nP = .001.\\n[SUBTITLE] Standardized incidence ratio [SUBSECTION] Thirty-five of the 66 counties had a SIRCOUNT\\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\\n was 2.15 and the lowest was 0.31.\\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\\nOn the other hand, for the SIRLM\\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\\n greater than 1. The 5 counties with the highest SIRLM\\n were similar to the SIRCOUNT\\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\\n was 1.94 and the lowest SIRLM\\n was 0.35. \\nThe results of the SIRs (SIRCOUNT\\n and SIRLM\\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\\n values. These counties with the highest SIRCOUNT\\n made up 5 of the 6 counties with the highest SIRLM\\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\\n values were the same 2 counties with the lowest SIRLM\\n values.\\nThirty-five of the 66 counties had a SIRCOUNT\\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\\n was 2.15 and the lowest was 0.31.\\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\\nOn the other hand, for the SIRLM\\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\\n greater than 1. The 5 counties with the highest SIRLM\\n were similar to the SIRCOUNT\\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\\n was 1.94 and the lowest SIRLM\\n was 0.35. \\nThe results of the SIRs (SIRCOUNT\\n and SIRLM\\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\\n values. These counties with the highest SIRCOUNT\\n made up 5 of the 6 counties with the highest SIRLM\\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\\n values were the same 2 counties with the lowest SIRLM\\n values.\",\n \"To learn more about the data, we performed an Exploratory Data Analysis (EDA) using several techniques. We explored the geographic distribution of the breast cancer mortality rates by using a choropleth map of South Dakota with age-adjusted mortality and incidence rates (Figure 1). The eastern side of the state had lower and more consistent mortality rates, followed by the far western part of the state. The central west part of the state exhibited higher mortality rates.\\nMap A shows the age-adjusted breast cancer mortality rates and Map B shows the age-adjusted breast cancer incidence rates, by county (N = 66), South Dakota, 2001–2015. Counties whose mortality rates have been imputed are marked with a star. Source: South Dakota State Cancer Registry, South Dakota Department of Health (12).\\nFive counties had higher mortality rates than the rest of the counties: Perkins, Mellette, Aurora, Douglas, and Corson. Two counties had lower mortality rates than the rest of the counties: Ziebach and Jackson. Eight counties (Bennet, Buffalo, Corson, Jackson, Mellette, Oglala, Todd, and Ziebach) had a poverty percentage greater than 30%, which was distinctly higher than the others; each county contains land on an American Indian Reservation. Minnehaha and Pennington counties had high screening rates; these counties are home to the first- and second-largest cities in South Dakota, respectively, so they also had the greatest number of screening providers.\\nAll variables, besides the number of providers and screening rate, had a correlation greater than zero with mortality rate. Incidence, educational attainment, and uninsured percentage all had a low positive correlation with mortality rate. The highest correlation with mortality rate was education attainment, at a value of 0.26. The remaining variables had a negligible correlation with mortality rate.\",\n \"Model 2 had the highest adjusted R\\n2 value (0.10) and the lowest AIC (441.79), with 2 significant factors associated with mortality rate; Model 3 had the lowest adjusted R\\n2 (.004) and the highest AIC (453.32), with no significant predictors of mortality. In Model 2, breast cancer incidence and educational attainment were predictors of breast cancer mortality, indicating that as more people are diagnosed with breast cancer and as the percentage of people with less than a bachelor’s degree increases, breast cancer mortality rate increases (Table 2). Educational attainment was a predictor of mortality in all models, and the educational attainment statistic had a P < .001. \\nAbbreviation: — , not applicable; AIC, Akaike information criterion.\\n\\nP = .01.\\n\\nP = .001.\",\n \"Thirty-five of the 66 counties had a SIRCOUNT\\n greater than 1, meaning that more than half of the counties had more deaths than expected (Figure 2). The 5 counties with the highest SIRCOUNT\\n, in decreasing order, were Perkins, Mellette, Aurora, Douglas, and Corson. Of those, Perkins, Mellette, and Aurora had more than twice the expected number of deaths. Ziebach, Jackson, Davison, Tripp, and Meade counties had the lowest SIRCOUNT\\n, with Ziebach and Jackson both being less than half the expected number of deaths. The highest SIRCOUNT\\n was 2.15 and the lowest was 0.31.\\nMap A shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment of the data, and Map B shows the predicted breast cancer mortality rate of South Dakota counties, accounting for age-adjustment, incidence rate, and educational attainment. Abbreviation: SIR, standardized incidence ratio. Map Sources: South Dakota State Cancer Registry, South Dakota Department of Health (12), Holzhauser et al (13), and the US Census Bureau (8).\\nOn the other hand, for the SIRLM\\n, most counties had fewer deaths than expected, with 28 of 66 counties having a SIRLM\\n greater than 1. The 5 counties with the highest SIRLM\\n were similar to the SIRCOUNT\\n: Corson, Perkins, Aurora, Jones, and Mellette in decreasing order. No county had more than twice the expected number of deaths. The 5 counties with the lowest SIRLM\\n were Ziebach, Jackson, Tripp, Oglala Lakota, and Davison. Ziebach and Jackson counties had less than half the number of expected deaths. The highest SIRLM\\n was 1.94 and the lowest SIRLM\\n was 0.35. \\nThe results of the SIRs (SIRCOUNT\\n and SIRLM\\n) are presented in Figure 2. The eastern side of the state showed similar SIR values while the western side of the state had more variation. Perkins, Mellette, Aurora, Douglas, and Corson counties had SIRCOUNT\\n values that were much higher than those of the rest of the counties. Ziebach and Jackson counties had the lowest SIRCOUNT\\n values. These counties with the highest SIRCOUNT\\n made up 5 of the 6 counties with the highest SIRLM\\n values, with Jones County replacing Douglas County. The 2 counties with the lowest SIRCOUNT\\n values were the same 2 counties with the lowest SIRLM\\n values.\",\n \"Overall, we found a significant association between incidence rate and educational level with respect to breast cancer mortality rates. Breast cancer incidence was positively associated with mortality rates in South Dakota, which suggests that more breast cancer cases are associated with more breast cancer deaths. In addition, educational attainment was repeatedly identified as a significant factor for mortality. Gadeyne et al found inconclusive results in their study of breast cancer mortality and education; however, Albano et al found a significant association between educational levels and cancer in general (7,25), specifically that lower educational attainment was related to higher cancer mortality rates, reflecting the findings of this study. Race, median age, and number of women screened were not selected in the feature selection during stepwise regression in our study; similarly, race, median age, and number of women screened were not significant in our full model.\\nThe 2020–2021 South Dakota Department of Education yearly review stated that American Indians were the largest minority group in school. However, American Indians still have a 63% completion rate for high school graduation and 59% attendance rate, compared with Whites who have a 94% completion rate for high school graduation and 94% attendance rate (26). An interesting point to consider is that South Dakota has no set standards for sex education (27). Thus, students are not taught reproductive health in general, including the importance of breast examinations, Pap smears, or prostate examinations. We advocate that set and scientifically backed health standards in high school would expose students at an early age to the risks of breast cancer and their options for screening.\\nThe western half of South Dakota had more variability in SIR values, and the state’s demographics could be a possible explanation. The 4 counties with the highest SIRs for both count and linear model SIRs were Corson, Perkins, Aurora, and Mellette, which are either in an American Indian reservation or neighbor a county within an American Indian reservation. Research on 3 tribes in western South Dakota supported that trust is often a barrier for American Indians (as are remote location and approvals by Indian Health Service programs) (28). Research in New Mexico reported that even after in-depth implementation of screening programs that lowered the barriers of cost, availability, and access to Native American and Hispanic women, the screening rates remained low, under 40% of women annually (26). The high SIRs in or neighboring reservation counties may mean that trust is also an issue, and South Dakota has more to work on than accessibility to Native Americans.\\nAfter controlling for incidence rate and educational attainment, the SIRLM\\n values became less variable. The SIRs’ decrease in range and mean closer to 1 indicate that the factors did affect mortality rate. This again agrees with findings from Albano et al that educational attainment affects mortality (7).\\nWe found that some counties had a higher mortality rate than expected based on the age of the women in the county. Ziebach and Jackson counties had the highest mortality rates, and the counties with the lowest SIRs are not home to major medical centers. Haakon County is vertically between Ziebach and Jackson counties and is one of the counties that does not have a provider; however, Haakon County has a higher mortality rate than the average of counties of South Dakota and both SIRs greater than 1, which means there were more deaths than expected. The areas of the map where there are dark green counties next to dark red counties are either on an American Indian reservation or neighbor an American Indian reservation. The differences between counties do not come from any singular cause, but rather due to variations in race, poverty levels, and population size.\\nOur study has limitations, primarily in the absence of portions of mortality data. Because South Dakota is largely a rural state, several counties have small populations and see very few deaths from breast cancer. These numbers are then withheld from the public to protect the privacy of the patients. This suppression resulted in having to impute the mortality rates of 15 counties, possibly introducing errors. The assumptions of the regression model did not account for this error, which may confer bias on the results. The counties with American Indian reservations have another health system that could have resulted in the under-representation or over-representation of breast cancer deaths in those counties (29). In addition, a study found that the misclassification of the race of Native Americans caused an underestimation of mortality rates as well (30). These gaps in databases and their contents highlight research challenges that rural communities will continue to face when few data are collected, populations and incidence are sparse, and data are inconsistently collected by multiple sources. Sociodemographic data are also challenging to consistently collect throughout a state. More detailed data per county would help yield accurate and unbiased results. For example, considering education, Zajacova and Lawrence argue that education is not a single-generation factor (31). Having data on the educational attainment of a patient’s parents or family, in addition to their own educational attainment, would allow us to assess the risk and see the relationship between education and breast cancer incidence or mortality. Thus, more research is needed to understand the effects education level has on financial security, stable employment, social success, and in turn, breast cancer mortality.\\nIn conclusion, understanding the risk factors and geographic distribution of breast cancer mortality among women across the state will assist stakeholders with efforts at prevention and resource allocation guided by data.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["other1","other2","other3","intro","methods",null,null,null,null,null,null,null,"results",null,null,null,"discussion"],"string":"[\n \"other1\",\n \"other2\",\n \"other3\",\n \"intro\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\"\n]"},"keywords":{"kind":"list like","value":[],"string":"[]"}}},{"rowIdx":365,"cells":{"title":{"kind":"string","value":"Effect of Comorbidities on the Infection Rate and Severity of COVID-19: Nationwide Cohort Study With Propensity Score Matching."},"pmid":{"kind":"string","value":"36265125"},"background_abstract":{"kind":"string","value":"A vaccine against COVID-19 has been developed; however, COVID-19 transmission continues. Although there have been many studies of comorbidities that have important roles in COVID-19, some studies have reported contradictory results."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Data were derived from a nationwide South Korean COVID-19 cohort study with propensity score (PS) matching. We included infected individuals who were COVID-19-positive between January 1, 2020, and May 30, 2020, and PS-matched uninfected controls. PS matching was performed to balance the baseline characteristics of each comorbidity and to adjust for potential confounders, such as age, sex, Charlson Comorbidity Index, medication, and other comorbidities, that were matched with binary variables. The outcomes were the confirmed comorbidities affecting the infection rate and severity of COVID-19. The endpoints were COVID-19 positivity and severe clinical outcomes of COVID-19 (such as tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, cardiopulmonary resuscitation, and death)."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The COVID-19 cohort with PS matching included 8070 individuals with positive COVID-19 test results and 8070 matched controls. The proportions of patients in the severe group were higher for individuals 60 years or older (severe clinical outcomes for those 60 years or older, 16.52%; severe clinical outcomes for those of other ages, 2.12%), those insured with Medicaid (Medicaid, 10.81%; other insurance, 5.61%), and those with disabilities (with disabilities, 18.26%; without disabilities, 5.07%). The COVID-19 infection rate was high for patients with pulmonary disease (odds ratio [OR] 1.88; 95% CI 1.70-2.03), dementia (OR 1.75; 95% CI 1.40-2.20), gastrointestinal disease (OR 1.74; 95% CI 1.62-1.88), stroke (OR 1.67; 95% CI 1.23-2.27), hepatobiliary disease (OR 1.31; 95% CI 1.19-1.44), diabetes mellitus (OR 1.28; 95% CI 1.16-1.43), and cardiovascular disease (OR 1.20; 95% CI 1.07-1.35). In contrast, it was lower for individuals with hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), and cancer (OR 0.73; 95% CI 0.62-0.86). The severity of COVID-19 was high for individuals with kidney disease (OR 5.59; 95% CI 2.48-12.63), hypertension (OR 2.92; 95% CI 1.91-4.47), dementia (OR 2.92; 95% CI 1.91-4.47), cancer (OR 1.84; 95% CI 1.15-2.94), pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), and psychotic disorders (OR 1.29; 95% CI 1.01-6.52). However, it was low for those with hyperlipidemia (OR 0.78; 95% CI 0.60-1.00)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Upon PS matching considering the use of statins, it was concluded that people with hyperlipidemia could have lower infection rates and disease severity of COVID-19."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["United States","Humans","Child, Preschool","COVID-19","SARS-CoV-2","Cohort Studies","Propensity Score","Cardiovascular Diseases","COVID-19 Vaccines","Comorbidity","Diabetes Mellitus","Hyperlipidemias","Dementia"],"string":"[\n \"United States\",\n \"Humans\",\n \"Child, Preschool\",\n \"COVID-19\",\n \"SARS-CoV-2\",\n \"Cohort Studies\",\n \"Propensity Score\",\n \"Cardiovascular Diseases\",\n \"COVID-19 Vaccines\",\n \"Comorbidity\",\n \"Diabetes Mellitus\",\n \"Hyperlipidemias\",\n \"Dementia\"\n]"},"pmcid":{"kind":"string","value":"9678330"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"The World Health Organization declared that COVID-19 was a pandemic in March 2020. By August 2022, approximately 600 million individuals had been infected, and more than 6 million had died. Since then, vaccines and therapeutic agents for COVID-19 have been developed. However, the current number of individuals with COVID-19 is still the same as that 1 year ago because it has not yet been eradicated [1]. COVID-19 can result in an asymptomatic presentation or flu-like symptoms. Some patients are admitted to the hospital for conservative treatment, and some require intensive care unit admission. Moreover, some patients may die as a result of COVID-19 [2,3]. As the number of individuals with COVID-19 increases, it is important to identify those who are vulnerable to severe COVID-19 to effectively manage health care resources accordingly and to improve the prognosis [4,5].\nSince the COVID-19 outbreak, many studies of the demographic factors that predispose individuals to infection and of the identification of comorbidities of infected individuals have been performed. Most studies have reported similar overall results; however, some results of these studies are contradictory [6-9]. These differences in results may be attributable to the diversity of patients and medical systems in various countries worldwide. Most previous studies on comorbidities analyzed the baseline characteristics of people infected with COVID-19 without considering the bias caused by various factors that influence COVID-19. For instance, to determine whether hyperlipidemia affects the severity of COVID-19, it is necessary to control for statins, which are often used by individuals with hyperlipidemia. Although some studies suggested that statins might have a role in reducing the severity of COVID-19 [10,11], most studies did not confirm the use of statins; they only reported the effect of hyperlipidemia [12-16]. Hence, it is difficult to accurately determine the effect of hyperlipidemia on the severity of COVID-19. We investigated the effects of patient comorbidities on the infection rate and severity of COVID-19. Bias was reduced by propensity score (PS) matching for various variables that may affect COVID-19. We also analyzed the demographic characteristics of patients with COVID-19."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Study Design and Participants [SUBSECTION] We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\nWe conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\n[SUBTITLE] Ethical Considerations [SUBSECTION] This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\nThis study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\n[SUBTITLE] Study Population [SUBSECTION] In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\nIn accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\n[SUBTITLE] Outcomes [SUBSECTION] To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\nTo determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\n[SUBTITLE] Statistical Analysis [SUBSECTION] We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\nWe performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\n[SUBTITLE] Patient and Public Involvement [SUBSECTION] No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\nNo patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Clinical Characteristics of the Study Population [SUBSECTION] A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nA total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\n[SUBTITLE] Risks of COVID-19 Positivity and Disease Severity According to Comorbidities [SUBSECTION] To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity.\nTo identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Study Design and Participants","Ethical Considerations","Study Population","Outcomes","Statistical Analysis","Patient and Public Involvement","Clinical Characteristics of the Study Population","Risks of COVID-19 Positivity and Disease Severity According to Comorbidities","Principal Findings","Limitations","Conclusions"],"string":"[\n \"Study Design and Participants\",\n \"Ethical Considerations\",\n \"Study Population\",\n \"Outcomes\",\n \"Statistical Analysis\",\n \"Patient and Public Involvement\",\n \"Clinical Characteristics of the Study Population\",\n \"Risks of COVID-19 Positivity and Disease Severity According to Comorbidities\",\n \"Principal Findings\",\n \"Limitations\",\n \"Conclusions\"\n]"},"other_sections_texts":{"kind":"list like","value":["We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.","This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.","In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.","To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.","We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).","No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.","A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).","To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity.","This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.","Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.","Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19."],"string":"[\n \"We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\",\n \"This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\",\n \"In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\",\n \"To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\",\n \"We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\",\n \"No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\",\n \"A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naParticipants from some specific groups, such as soldiers, were not included.\\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\",\n \"To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\\nbComorbidity with more susceptibility to COVID-19.\\ncComorbidity with less susceptibility to COVID-19.\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\\nbComorbidity with increasing COVID-19 severity.\\ncComorbidity with decreasing COVID-19 severity.\",\n \"This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\",\n \"Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\",\n \"Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Study Design and Participants","Ethical Considerations","Study Population","Outcomes","Statistical Analysis","Patient and Public Involvement","Results","Clinical Characteristics of the Study Population","Risks of COVID-19 Positivity and Disease Severity According to Comorbidities","Discussion","Principal Findings","Limitations","Conclusions"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Study Design and Participants\",\n \"Ethical Considerations\",\n \"Study Population\",\n \"Outcomes\",\n \"Statistical Analysis\",\n \"Patient and Public Involvement\",\n \"Results\",\n \"Clinical Characteristics of the Study Population\",\n \"Risks of COVID-19 Positivity and Disease Severity According to Comorbidities\",\n \"Discussion\",\n \"Principal Findings\",\n \"Limitations\",\n \"Conclusions\"\n]"},"all_sections_texts":{"kind":"list like","value":["The World Health Organization declared that COVID-19 was a pandemic in March 2020. By August 2022, approximately 600 million individuals had been infected, and more than 6 million had died. Since then, vaccines and therapeutic agents for COVID-19 have been developed. However, the current number of individuals with COVID-19 is still the same as that 1 year ago because it has not yet been eradicated [1]. COVID-19 can result in an asymptomatic presentation or flu-like symptoms. Some patients are admitted to the hospital for conservative treatment, and some require intensive care unit admission. Moreover, some patients may die as a result of COVID-19 [2,3]. As the number of individuals with COVID-19 increases, it is important to identify those who are vulnerable to severe COVID-19 to effectively manage health care resources accordingly and to improve the prognosis [4,5].\nSince the COVID-19 outbreak, many studies of the demographic factors that predispose individuals to infection and of the identification of comorbidities of infected individuals have been performed. Most studies have reported similar overall results; however, some results of these studies are contradictory [6-9]. These differences in results may be attributable to the diversity of patients and medical systems in various countries worldwide. Most previous studies on comorbidities analyzed the baseline characteristics of people infected with COVID-19 without considering the bias caused by various factors that influence COVID-19. For instance, to determine whether hyperlipidemia affects the severity of COVID-19, it is necessary to control for statins, which are often used by individuals with hyperlipidemia. Although some studies suggested that statins might have a role in reducing the severity of COVID-19 [10,11], most studies did not confirm the use of statins; they only reported the effect of hyperlipidemia [12-16]. Hence, it is difficult to accurately determine the effect of hyperlipidemia on the severity of COVID-19. We investigated the effects of patient comorbidities on the infection rate and severity of COVID-19. Bias was reduced by propensity score (PS) matching for various variables that may affect COVID-19. We also analyzed the demographic characteristics of patients with COVID-19.","[SUBTITLE] Study Design and Participants [SUBSECTION] We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\nWe conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\n[SUBTITLE] Ethical Considerations [SUBSECTION] This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\nThis study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\n[SUBTITLE] Study Population [SUBSECTION] In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\nIn accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\n[SUBTITLE] Outcomes [SUBSECTION] To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\nTo determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\n[SUBTITLE] Statistical Analysis [SUBSECTION] We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\nWe performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\n[SUBTITLE] Patient and Public Involvement [SUBSECTION] No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\nNo patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.","We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.","This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.","In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.","To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.","We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).","No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.","[SUBTITLE] Clinical Characteristics of the Study Population [SUBSECTION] A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nA total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\n[SUBTITLE] Risks of COVID-19 Positivity and Disease Severity According to Comorbidities [SUBSECTION] To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity.\nTo identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity.","A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naParticipants from some specific groups, such as soldiers, were not included.\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).","To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\nbComorbidity with more susceptibility to COVID-19.\ncComorbidity with less susceptibility to COVID-19.\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\nbComorbidity with increasing COVID-19 severity.\ncComorbidity with decreasing COVID-19 severity.","[SUBTITLE] Principal Findings [SUBSECTION] This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\nThis study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\n[SUBTITLE] Limitations [SUBSECTION] Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\nOur study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\n[SUBTITLE] Conclusions [SUBSECTION] Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.\nAlthough the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.","This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.","Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.","Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19."],"string":"[\n \"The World Health Organization declared that COVID-19 was a pandemic in March 2020. By August 2022, approximately 600 million individuals had been infected, and more than 6 million had died. Since then, vaccines and therapeutic agents for COVID-19 have been developed. However, the current number of individuals with COVID-19 is still the same as that 1 year ago because it has not yet been eradicated [1]. COVID-19 can result in an asymptomatic presentation or flu-like symptoms. Some patients are admitted to the hospital for conservative treatment, and some require intensive care unit admission. Moreover, some patients may die as a result of COVID-19 [2,3]. As the number of individuals with COVID-19 increases, it is important to identify those who are vulnerable to severe COVID-19 to effectively manage health care resources accordingly and to improve the prognosis [4,5].\\nSince the COVID-19 outbreak, many studies of the demographic factors that predispose individuals to infection and of the identification of comorbidities of infected individuals have been performed. Most studies have reported similar overall results; however, some results of these studies are contradictory [6-9]. These differences in results may be attributable to the diversity of patients and medical systems in various countries worldwide. Most previous studies on comorbidities analyzed the baseline characteristics of people infected with COVID-19 without considering the bias caused by various factors that influence COVID-19. For instance, to determine whether hyperlipidemia affects the severity of COVID-19, it is necessary to control for statins, which are often used by individuals with hyperlipidemia. Although some studies suggested that statins might have a role in reducing the severity of COVID-19 [10,11], most studies did not confirm the use of statins; they only reported the effect of hyperlipidemia [12-16]. Hence, it is difficult to accurately determine the effect of hyperlipidemia on the severity of COVID-19. We investigated the effects of patient comorbidities on the infection rate and severity of COVID-19. Bias was reduced by propensity score (PS) matching for various variables that may affect COVID-19. We also analyzed the demographic characteristics of patients with COVID-19.\",\n \"[SUBTITLE] Study Design and Participants [SUBSECTION] We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\\nWe conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\\n[SUBTITLE] Ethical Considerations [SUBSECTION] This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\\nThis study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\\n[SUBTITLE] Study Population [SUBSECTION] In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\\nIn accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\\n[SUBTITLE] Outcomes [SUBSECTION] To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\\nTo determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\\n[SUBTITLE] Statistical Analysis [SUBSECTION] We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\\nWe performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\\n[SUBTITLE] Patient and Public Involvement [SUBSECTION] No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\\nNo patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\",\n \"We conducted a large-scale cohort study using a South Korean National Health Insurance claims database [17]. In South Korea, all citizens are registered in the Korean National Health Insurance Service (KNHIS) database. The KNHIS uses a nationwide, large-scale database system including information regarding the diagnostic codes from the International Classification of Diseases (ICD)-10, the names of the procedures performed, prescription drugs, hospital information, direct medical costs of inpatient and outpatient treatments, and medical insurance premiums. Because all Koreans are given unique identification numbers at birth that are used in the KNHIS, the health records of patients are not duplicated nor omitted [18,19]. For COVID-19 studies, KNHIS provides a COVID-19 cohort that includes people infected with COVID-19 and a control group that had never been infected. From January 1, 2020, to May 31, 2020, disease codes B342, B972, U071, U072, MT043, and 3/02 were used to identify patients with confirmed COVID-19. Data from the control group of individuals who were not previously diagnosed with COVID-19 were adjusted for sex, age, and region of residence. Moreover, the number of participants in the control group was 15 times the number of confirmed COVID-19 cases.\",\n \"This study was approved by the relevant institutional review board and research ethics committee (ISPAIK 2020-06-048-001). The need for written consent was formally waived by the ethics committee. This study used the NHIS-2020-1-328 database provided by the KNHIS in 2020.\",\n \"In accordance with the World Health Organization guidelines, laboratory confirmation of COVID-19 was defined as a positive result of a real-time reverse-transcription polymerase chain reaction assay using a sample obtained with nasal and pharyngeal swabs [20]. We combined the claims-based data from the KNHIS between January 1, 2015, and May 31, 2020, and extracted information regarding age, sex, and region of residence from the insurance eligibility data (Figure 1). The Charlson Comorbidity Index (CCI) score was calculated using the ICD-10 codes and previously reported methods [21]. Certain underlying medications and diseases with a high risk of serious illness attributable to SARS-CoV-2, which causes COVID-19, were studied and reported by the Centers for Disease Control and Prevention (CDC) and previous meta-analysis studies [6-9,22]. In these studies, we selected factors to use for PS matching in the analysis (Tables S1 and S2 in Multimedia Appendix 1). Only those (pulmonary disease, cardiovascular disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, and psychotic disorder) with more than 500 people with COVID-19 were selected because a small number of people with corresponding comorbidities might cause statistical bias (Multimedia Appendix 2). A history of underlying diseases (pulmonary disease, cardiovascular disease, kidney disease, hepatobiliary disease, hyperlipidemia, gastrointestinal disease, diabetes mellitus, hypertension, psychotic disorder, dementia, stroke, neurologic disorder, autoimmune disease, and cancer) was confirmed by the assignment of at least two claims within 1 year using the appropriate ICD-10 code.\\nWe used various PS matching methods for factors affecting COVID-19: (1) matching for age, sex, and CCI; (2) additional matching for comorbidities; and (3) additional matching for medications. Finally, the results from (3) were used (Tables S3 and S4 in Multimedia Appendix 1). The financial revenue of the National Health Insurance of Korea consists of contributions from the insured and government subsidies, which can be used to analyze socioeconomic status. The contributions to the National Health Insurance differ according to the family income level. The higher the income, the greater the contribution to the National Health Insurance. Income was divided into 5 categories for the purpose of statistical analyses. The first category is Medicaid, and the successive categories include progressively higher (by 25%) income groups. Disability grades were categorized as mild or severe based on the KNHIS database information for people registered with the Korean government.\\nDisposition of patients in the KNHIS-COVID cohort (South Korea; January 1 to May 31, 2020). CCI: Charlson Comorbidity Index; KNHIS: Korean National Health Insurance Service.\",\n \"To determine the severity of disease according to the demographic factors of COVID-19–infected patients, the severity scale was divided into the following 4 grades: mild, moderate, severe, and death. In South Korea, patients with asymptomatic or mild symptoms are discharged when a negative COVID-19 test result is confirmed 2 weeks after hospitalization. This time period also corresponds to the period of self-isolation. When we checked the hospitalization period of COVID-19–infected patients, the hospitalization period peaked on day 16 and decreased thereafter. Based on this result, a hospitalization period of ≤16 days was defined as the mild grade corresponding to asymptomatic or mild symptoms. The severe grade was defined as the need for tracheostomy, continuous renal replacement therapy, intensive care unit admission, ventilator use, and cardiopulmonary resuscitation. The moderate grade was defined as a hospitalization period >16 days but not requiring treatment corresponding to the severe grade.\\nThe primary aim of this study was to compare the severity grades of the COVID-19–infected and control groups based on demographic factors, comorbidities, and complications. The secondary aim was to perform PS matching for comparisons. We identified the infection rate and severity (severe and death or mild and moderate) of COVID-19 according to the comorbid conditions.\",\n \"We performed PS matching to balance the baseline characteristics of each comorbidity (existence or nonexistence) and to adjust for potential confounders. Because we focused on each comorbidity, PS matching was performed for each comorbidity. The PS was estimated using a logistic regression model and calculating the predicted probability of covariates. Age and CCI (0, 1, or ≥2) were matched with continuous variables. Sex, medication, and other comorbidities were matched with binary variables. We assessed the PS matching of the comorbidity existence using a 1:1 ratio, the greedy nearest neighbor algorithm, and a scale with a caliper of 0.25 (Multimedia Appendix 2). Data obtained after PS matching were analyzed by calculating the odds ratios (ORs) with 95% CIs for the infection rate and severity (severe and death or mild and moderate) of COVID-19. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).\",\n \"No patient was directly involved in designing the study question or in conducting the study. No patients were asked for advice regarding the interpretation or writing of the results. There are no plans to involve patients or the relevant patient community in the dissemination of study findings at this time.\",\n \"[SUBTITLE] Clinical Characteristics of the Study Population [SUBSECTION] A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naParticipants from some specific groups, such as soldiers, were not included.\\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\nA total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naParticipants from some specific groups, such as soldiers, were not included.\\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\n[SUBTITLE] Risks of COVID-19 Positivity and Disease Severity According to Comorbidities [SUBSECTION] To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\\nbComorbidity with more susceptibility to COVID-19.\\ncComorbidity with less susceptibility to COVID-19.\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\\nbComorbidity with increasing COVID-19 severity.\\ncComorbidity with decreasing COVID-19 severity.\\nTo identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\\nbComorbidity with more susceptibility to COVID-19.\\ncComorbidity with less susceptibility to COVID-19.\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\\nbComorbidity with increasing COVID-19 severity.\\ncComorbidity with decreasing COVID-19 severity.\",\n \"A total of 8070 individuals had positive COVID-19 results according to the reverse-transcription polymerase chain reaction assay. We identified 121,050 uninfected individuals as control participants (Multimedia Appendix 2). The demographic characteristics of the entire cohort are displayed in Table 1. The COVID-19 severity grade was mild for 2419 (2419/8070, 29.98%) individuals, moderate for 5160 (5160/8070, 63.94%) individuals, severe for 254 (254/8070, 3.15%) individuals, and death for 237 (237/8070, 2.94%) individuals. Among the total sample of infected individuals, 3236 (3236/8070, 40.10%) were male. Most patients were in their fifth (1567/8070, 19.42%) or sixth (1199/8070, 14.86%) decade of life. In terms of the medical insurance grade, which indicates socioeconomic status, those receiving Medicaid had high rates of severe grade and death. However, there were no obvious trends for the other grades. Individuals with disabilities had more severe infections and a much higher case fatality rate (Table 1). Those with COVID-19 had a medical history of gastrointestinal disease (n=5256), pulmonary disease (n=2539), hyperlipidemia (n=1841), and hypertension (n=1623). The case fatality rate was high for individuals with dementia (74/235, 31.5%), kidney disease (25/86, 29%), and cardiovascular disease (110/675, 16.3%; Table 2). After COVID-19 was confirmed, gastrointestinal disease (n=2912), pulmonary disease (n=2398), and hepatobiliary disease (n=1248) were the most common complications (Table 3).\\nBaseline characteristics of the study population, including those infected (n=8070) and uninfected (n=121,050; controls) with COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naParticipants from some specific groups, such as soldiers, were not included.\\nbThe uninfected controls were adjusted for sex, age, and region, resulting in a figure equivalent to 15 times the number of confirmed COVID-19 cases in the KNHIS-COVID cohort.\\nBaseline characteristics of comorbidities of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\nBaseline characteristics of complications of the study population, including those infected with (n=8070) and not infected with (n=121,050; controls) COVID-19 in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\",\n \"To identify differences according to comorbidity, predispositions were matched between the COVID-19–infected group and uninfected control group. No significant imbalances in the demographics and clinical characteristics were observed when they were assessed using the standardized mean difference within groups of PS-matched cohorts, which included the standardized mean difference of binary type variables <0.1. PS-matched ORs were checked for age, sex, CCI, medication, and comorbidities. When the control group and COVID-19–infected group were compared, COVID-19 was likely to occur in individuals with a history of the diseases and medical conditions but not for those with a history of hyperlipidemia (OR 0.73; 95% CI 0.67-0.80), autoimmune disease (OR 0.73; 95% CI 0.60-0.89), or cancer (OR 0.73; 95% CI 0.62-0.86; Table 4). The severity grade was high for COVID-19–infected individuals with pulmonary disease (OR 1.72; 95% CI 1.35-2.19), cardiovascular disease (OR 1.54; 95% CI 1.17-2.04), kidney disease (OR 5.59; 95% CI 2.48-12.63), diabetes mellitus (OR 1.43; 95% CI 1.09-1.87), hypertension (OR 1.63; 95% CI 1.23-2.15), psychotic disorder (OR 1.29; 95% CI 1.01-6.52), dementia (OR 2.92; 95% CI 1.91-4.47), or cancer (OR 1.84; 95% CI 1.15-2.94). However, the severity grade was low for COVID-19–infected individuals with hyperlipidemia (OR 0.70; 95% CI 0.55-0.90; Table 5 and Table S5 in Multimedia Appendix 1).\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and COVID-19 infection positivity rates according to comorbidity in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the COVID-19 and control groups.\\nbComorbidity with more susceptibility to COVID-19.\\ncComorbidity with less susceptibility to COVID-19.\\nPropensity score–matched (age, sex, Charlson Comorbidity Index, medications, and comorbidities) baseline characteristics and clinical outcomes of COVID-19 among patients in the mild or moderate group and those in the severe or death group according to the comorbidity of patients with laboratory-confirmed COVID-19 infection in the Korean National Health Insurance Service (KNHIS)-COVID cohort (South Korea; January 1, 2020, to May 31, 2020).\\naWe assessed each propensity score–matched comorbidity using a 1:1 ratio for those in the mild and moderate group and those in the severe and death group.\\nbComorbidity with increasing COVID-19 severity.\\ncComorbidity with decreasing COVID-19 severity.\",\n \"[SUBTITLE] Principal Findings [SUBSECTION] This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\\nThis study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\\n[SUBTITLE] Limitations [SUBSECTION] Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\\nOur study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\\n[SUBTITLE] Conclusions [SUBSECTION] Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.\\nAlthough the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.\",\n \"This study was a retrospective cohort study conducted in South Korea from January 2020 to May 2020. It involved confirmed COVID-19 patients with medical insurance. Previous studies of the demographic factors of individuals with COVID-19 showed that male sex, old age, and low income were factors likely associated with COVID-19 with a high severity grade [23,24]. In this study, more women had COVID-19, but the severity grade of COVID-19 was higher for men; this was directly proportional to age, especially for men older than 70 years. All medical expenses for COVID-19 are paid for by the South Korea government; therefore, all patients, including those receiving Medicaid, received the same level of care for COVID-19. Although there was no difference in medical care, those with Medicaid had the lowest income level and a higher severity grade; however, there were no differences between the groups with grades 1 to 4 medical insurance. For individuals with disabilities, the incidence was slightly higher than that of the control group. However, the severity grade was much higher than that of other individuals infected with COVID-19.\\nOther studies of COVID-19 reported that SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptor through the viral structural spike protein at the onset of infection [25]. ACE2 is expressed to varying degrees in almost all human organs. ACE2 is highly expressed in cardiomyocytes, proximal tubule cells of the kidney, and bladder urinary tract cells. Additionally, it is abundantly expressed in intestinal cells of the small intestine, especially in the ileum [25-28]. Therefore, most critically ill patients with COVID-19 experience multiple organ injuries, including acute lung injuries, acute kidney injuries, cardiac injuries, hepatobiliary disease, and pneumothorax [29]. Therefore, to analyze the effect of each comorbidity on the COVID-19 infection severity grade, it is necessary to consider other comorbidities.\\nEach demographic factor, comorbidity, and medication may influence each other, resulting in different outcomes in terms of the infection rate and severity of COVID-19. When analyzing comorbidities with hypertension, the effect of hypertension on the infection rate and severity of COVID-19 experienced by an 80-year-old woman with asthma and that of a 30-year-old man without an underlying medical condition may be different. Accurate results can be obtained for sufficiently studied diseases by controlling for only important factors. However, in the case of understudied diseases, such as COVID-19, various factors should be considered. In this study, PS matching was performed for various factors that could affect COVID-19, to minimize bias. When selecting a factor for PS matching, in order to select objective data, data provided by the CDC and meta-analysis studies were used. However, there was a limit, as data may change as research on COVID-19 progresses. Most of the results obtained were similar to those of previously published studies; however, some results were conflicting. For people with cancer and autoimmune disease, infection rates were even lower; these results were possibly affected by reducing social contact because of the risk of COVID-19 infection. Exposure to COVID-19 is an important factor that can affect the infection rate of COVID-19. Individuals with hyperlipidemia had a low COVID-19 infection rate and low severity grade. Previous studies reported that hyperlipidemia should be managed to prevent COVID-19 because high cholesterol levels induce inflammation and increase ACE2 availability [30-32]. Moreover, the use of statins for patients with COVID-19 reduced mortality by interfering with the mevalonate pathway and because of their antiviral effects [10,11,33]. However, some studies have shown that people with low lipid levels are more susceptible to and have more severe COVID-19 infection [34-41]. A meta-analysis published in 2022 indicated that patients with severe COVID-19 had lower total cholesterol levels (pooled mean difference –10.4; 95% CI –18.7 to –2.2), low-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –8.4 to –0.42), and high-density lipoprotein cholesterol levels (pooled mean difference –4.4; 95% CI –6.9 to –1.8) on admission compared with patients with non-severe disease [42]. This may be similar to the “obesity paradox,” which states that mild obesity is advantageous to improvements after stroke [43,44]. Mild obesity can withstand the systemic catabolic imbalance with impaired metabolic efficiency and body tissue degradation that occur after stroke. Hyperlipidemia may also have a role in minimizing the severity of COVID-19.\",\n \"Our study has several limitations. As a limitation of most medical data studies, there is bias caused by confounding factors that may affect our results. When selecting factors for PS matching, information from the CDC and meta-analysis studies were used to select objective data, but these data may change as research on COVID-19 progresses. Further, we defined diseases based on the ICD codes provided in the insurance claims data. There may have been additional unmeasured confounders influencing our results, including genetic polymorphisms, smoking, body mass index, and exposure to the virus. In this study, the infection rate of COVID-19 may have been influenced by the degree of exposure to COVID-19, which may be an important factor in addition to the comorbidity factors. However, the influence of COVID-19 itself could be confirmed because the bias was less than that of previous studies. One race in South Korea comprises more than 95% of the population; hence, there was minimal racial bias compared with previous studies. Because the government funds the treatment for COVID-19 in South Korea and because the medical facilities for COVID-19 treatment are ubiquitous, there was minimal economic bias. The PS matching was performed for sex, age, CCI, comorbidity, and medication, including statins (standardized mean difference <0.1). Hence, selection bias was minimized. Therefore, more accurate information regarding the incidence of COVID-19 and its severity according to comorbidities was provided.\\nThis study was based on data from patients who experienced COVID-19 during the early outbreak period; therefore, that strain may differ from the current strain of COVID-19. However, an accurate analysis of recent COVID-19 strains, including Omicron, is difficult because the effects of acquired or natural immunity and vaccination are mixed. Data at the time of its early onset can provide fundamental information, including regarding mutations that may occur in the future.\",\n \"Although the severity of COVID-19 has decreased, its hospitalization rate has not decreased significantly, and its burden on medical facilities continues; therefore, an analysis of comorbidities is still important. Therefore, many studies of comorbidities that affect COVID-19 have been published; however, some have reported conflicting results. This may be because various factors such as medication and comorbidities, in addition to demographic factors such as age and sex, affect the infection rate and severity of COVID-19. It is necessary to analyze as many factors as possible to obtain more accurate data regarding COVID-19. Based on the results of previous studies, this study tried to derive objective results by considering various factors affecting COVID-19. In conclusion, certain comorbidities known as risk factors in previous studies increase the infection rate and severity of COVID-19. However, hyperlipidemia decreases the infection rate and severity. These results can be utilized to effectively manage COVID-19.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction","methods",null,null,null,null,null,null,"results",null,null,"discussion",null,null,null],"string":"[\n \"introduction\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n \"discussion\",\n null,\n null,\n null\n]"},"keywords":{"kind":"list like","value":["COVID-19","comorbidity","infection rate","severity of illness index","hyperlipidemia"],"string":"[\n \"COVID-19\",\n \"comorbidity\",\n \"infection rate\",\n \"severity of illness index\",\n \"hyperlipidemia\"\n]"}}},{"rowIdx":366,"cells":{"title":{"kind":"string","value":"New Surveillance Metrics for Alerting Community-Acquired Outbreaks of Emerging SARS-CoV-2 Variants Using Imported Case Data: Bayesian Markov Chain Monte Carlo Approach."},"pmid":{"kind":"string","value":"36265134"},"background_abstract":{"kind":"string","value":"Global transmission from imported cases to domestic cluster infections is often the origin of local community-acquired outbreaks when facing emerging SARS-CoV-2 variants."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We used Taiwanese COVID-19 weekly data on imported cases, domestic cluster infections, and community-acquired outbreaks. The study period included the D614G strain in February 2020, the Alpha and Delta variants of concern (VOCs) in 2021, and the Omicron BA.1 and BA.2 VOCs in April 2022. The number of cases arising from domestic cluster infection caused by imported cases (Dci/Imc) per week was used as the SARS-CoV-2 strain-dependent surveillance metric for alerting local community-acquired outbreaks. Its upper 95% credible interval was used as the alert threshold for guiding the rapid preparedness of containment measures, including nonpharmaceutical interventions (NPIs), testing, and vaccination. The 2 metrics were estimated by using the Bayesian Monte Carlo Markov Chain method underpinning the directed acyclic graphic diagram constructed by the extra-Poisson (random-effect) regression model. The proposed model was also used to assess the most likely week lag of imported cases prior to the current week of domestic cluster infections."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"A 1-week lag of imported cases prior to the current week of domestic cluster infections was considered optimal. Both metrics of Dci/Imc and the alert threshold varied with SARS-CoV-2 variants and available containment measures. The estimates were 9.54% and 12.59%, respectively, for D614G and increased to 14.14% and 25.10%, respectively, for the Alpha VOC when only NPIs and testing were available. The corresponding figures were 10.01% and 13.32% for the Delta VOC, but reduced to 4.29% and 5.19% for the Omicron VOC when NPIs, testing, and vaccination were available. The rapid preparedness of containment measures guided by the estimated metrics accounted for the lack of community-acquired outbreaks during the D614G period, the early Alpha VOC period, the Delta VOC period, and the Omicron VOC period between BA.1 and BA.2. In contrast, community-acquired outbreaks of the Alpha VOC in mid-May 2021, Omicron BA.1 VOC in January 2022, and Omicron BA.2 VOC from April 2022 onwards, were indicative of the failure to prepare containment measures guided by the alert threshold."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"We developed new surveillance metrics for estimating the risk of domestic cluster infections with increasing imported cases and its alert threshold for community-acquired infections varying with emerging SARS-CoV-2 strains and the availability of containment measures. The use of new surveillance metrics is important in the rapid preparedness of containment measures for averting large-scale community-acquired outbreaks arising from emerging imported SARS-CoV-2 variants."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","SARS-CoV-2","Markov Chains","Bayes Theorem","Benchmarking","COVID-19","Disease Outbreaks"],"string":"[\n \"Humans\",\n \"SARS-CoV-2\",\n \"Markov Chains\",\n \"Bayes Theorem\",\n \"Benchmarking\",\n \"COVID-19\",\n \"Disease Outbreaks\"\n]"},"pmcid":{"kind":"string","value":"9746786"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"During the COVID-19 pandemic lasting for over 2.5 years, countries around the world have experienced cyclical COVID-19 changes alternating between lifting and operating nonpharmaceutical interventions (NPIs) and between the protective and waning effects of vaccines when facing the incessant epidemics of the COVID-19 pandemic [1-4]. The cyclical resurgence of COVID-19 at the country, continental, and global levels is mainly caused by emerging SARS-CoV-2 variants, particularly variants of concern (VOCs). In response to the resurgence of community-acquired outbreaks, 2 containment measures have become important, including the timely adjustment of NPIs (strengthened border control strategies and restricted social activities) combined with testing and the launch of mass primary and booster vaccinations [5-7].\nIt should be noted that the typical pattern of transmission from an imported case to domestic cluster infection is often the root of local community-acquired outbreaks caused by emerging SARS-CoV-2 variants from any region or country across the globe [2,7-9]. Such an importation-cluster transmission mode has been clearly demonstrated by the resurgence of global epidemic waves following the emergence of dominant strains of Alpha, Beta, Gamma, Delta, and Omicron VOCs. To avert local community-acquired outbreaks of emerging SARS-CoV-2 variants, rapid preparedness of containment measures and effective contact tracing are mandatory when domestic cluster infections are identified after the introduction of emerging imported cases. In addition, the risk of domestic cluster infection on the introduction of imported cases varies with each emerging SARS-CoV-2 strain owing to the evolutionary characteristics of invading VOCs, including an increase in transmissibility and a higher likelihood of escaping immune response after vaccination [4,7-12].\nIt is therefore important to have new surveillance metrics for monitoring the odds of having domestic cluster infection transmitted from few imported cases and setting up the alert threshold for forestalling community-acquired outbreaks, as traditional surveillance metrics, like effective reproductive number (Rt), are tailored for assessing the spread and control of community-acquired outbreaks at the population level, which may only involve a single country and a specific SARS-CoV-2 strain in a short period rather than the country of the imported case across the world and the full spectrum of SARS-CoV-2 strains with a long period [13-16]. Such a traditional epidemic surveillance model (eg, the SEIR [Susceptible-Exposed-Infected-Recovery] model) is not only limited to model the relationship of sparse cases of domestic cluster infection and small samples of imported cases from each original country, but also inflexible to make allowance for the heterogeneity of the imported-domestic transmission mode across countries and SARS-CoV-2 strains across time, as well as the variation across local regions in question. To consider these issues of heterogeneity, it is therefore necessary to develop new surveillance models and their corresponding metrics with a new statistical approach, such as a sampling method of machine learning, particularly the Bayesian Markov Chain Monte Carlo (MCMC) method, in conjunction with a sparse event history regression model, such as the extra-Poisson (random-effect) regression model with relevant parameters and random variables parameterized under the directed acyclic graphic (DAG) diagram.\nDeveloping these new surveillance metrics for quantifying the effect size of the transmission from importation to domestic cluster infection would not only be helpful for alerting emerging community-acquired outbreaks, but also aid health professionals having rapid preparedness of SARS-CoV-2 strain–dependent containment measures, including effective and efficient contact tracing. Using a series of chronological epidemic data on COVID-19 divided into 2 phases (non-VOC phase [wild type and D614G] and VOC phase) in Taiwan, this study aimed to develop new surveillance metrices across the periods of various SARS-CoV-2 strains for alerting emerging community-acquired outbreaks by monitoring the risk of small domestic cluster infections originating from the transmission of few imported cases of emerging variants in order to forestall community-acquired outbreaks when facing emerging SARS-CoV-2 variants. The Bayesian MCMC sampling method was therefore used to estimate and predict the new surveillance metrics underpinning the DAG diagram of the Poisson or negative binomial random-effect regression model."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Data Sources [SUBSECTION] Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\nPublicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\n[SUBTITLE] Containment Measures for the Non-VOC Phase in Taiwan [SUBSECTION] The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\nThe containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\n[SUBTITLE] Containment Measures for the VOC Phase in Taiwan [SUBSECTION] The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\nThe Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\n[SUBTITLE] Statistical Analysis [SUBSECTION] [SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\n[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc)."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Evaluation of the Optimal Time Lag Model [SUBSECTION] After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\nAfter the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\n[SUBTITLE] Surveillance Metrics for the Imported-Domestic Transmission Mode [SUBSECTION] The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nThe weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\n[SUBTITLE] External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand [SUBSECTION] To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\nTo validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Data Sources","Containment Measures for the Non-VOC Phase in Taiwan","Containment Measures for the VOC Phase in Taiwan","Statistical Analysis","New Surveillance Metrics for Quantifying Imported-Domestic Transmission","Estimation With the Bayesian MCMC Method","Evaluation of the Optimal Time Lag Model","Surveillance Metrics for the Imported-Domestic Transmission Mode","Wild-Type and D614G Period","Alpha VOC Period","Delta and Omicron VOC Period","External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand"],"string":"[\n \"Data Sources\",\n \"Containment Measures for the Non-VOC Phase in Taiwan\",\n \"Containment Measures for the VOC Phase in Taiwan\",\n \"Statistical Analysis\",\n \"New Surveillance Metrics for Quantifying Imported-Domestic Transmission\",\n \"Estimation With the Bayesian MCMC Method\",\n \"Evaluation of the Optimal Time Lag Model\",\n \"Surveillance Metrics for the Imported-Domestic Transmission Mode\",\n \"Wild-Type and D614G Period\",\n \"Alpha VOC Period\",\n \"Delta and Omicron VOC Period\",\n \"External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand\"\n]"},"other_sections_texts":{"kind":"list like","value":["Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].","The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.","The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.","[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).","We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).","The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).","After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).","The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.","During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.","There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.","After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.","To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8."],"string":"[\n \"Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\",\n \"The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\",\n \"The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\",\n \"[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\",\n \"We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\",\n \"The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\",\n \"After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\",\n \"The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\\naVOC: variant of concern.\\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\",\n \"During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\",\n \"There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\",\n \"After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\",\n \"To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Data Sources","Containment Measures for the Non-VOC Phase in Taiwan","Containment Measures for the VOC Phase in Taiwan","Statistical Analysis","New Surveillance Metrics for Quantifying Imported-Domestic Transmission","Estimation With the Bayesian MCMC Method","Results","Evaluation of the Optimal Time Lag Model","Surveillance Metrics for the Imported-Domestic Transmission Mode","Wild-Type and D614G Period","Alpha VOC Period","Delta and Omicron VOC Period","External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand","Discussion"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Data Sources\",\n \"Containment Measures for the Non-VOC Phase in Taiwan\",\n \"Containment Measures for the VOC Phase in Taiwan\",\n \"Statistical Analysis\",\n \"New Surveillance Metrics for Quantifying Imported-Domestic Transmission\",\n \"Estimation With the Bayesian MCMC Method\",\n \"Results\",\n \"Evaluation of the Optimal Time Lag Model\",\n \"Surveillance Metrics for the Imported-Domestic Transmission Mode\",\n \"Wild-Type and D614G Period\",\n \"Alpha VOC Period\",\n \"Delta and Omicron VOC Period\",\n \"External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand\",\n \"Discussion\"\n]"},"all_sections_texts":{"kind":"list like","value":["During the COVID-19 pandemic lasting for over 2.5 years, countries around the world have experienced cyclical COVID-19 changes alternating between lifting and operating nonpharmaceutical interventions (NPIs) and between the protective and waning effects of vaccines when facing the incessant epidemics of the COVID-19 pandemic [1-4]. The cyclical resurgence of COVID-19 at the country, continental, and global levels is mainly caused by emerging SARS-CoV-2 variants, particularly variants of concern (VOCs). In response to the resurgence of community-acquired outbreaks, 2 containment measures have become important, including the timely adjustment of NPIs (strengthened border control strategies and restricted social activities) combined with testing and the launch of mass primary and booster vaccinations [5-7].\nIt should be noted that the typical pattern of transmission from an imported case to domestic cluster infection is often the root of local community-acquired outbreaks caused by emerging SARS-CoV-2 variants from any region or country across the globe [2,7-9]. Such an importation-cluster transmission mode has been clearly demonstrated by the resurgence of global epidemic waves following the emergence of dominant strains of Alpha, Beta, Gamma, Delta, and Omicron VOCs. To avert local community-acquired outbreaks of emerging SARS-CoV-2 variants, rapid preparedness of containment measures and effective contact tracing are mandatory when domestic cluster infections are identified after the introduction of emerging imported cases. In addition, the risk of domestic cluster infection on the introduction of imported cases varies with each emerging SARS-CoV-2 strain owing to the evolutionary characteristics of invading VOCs, including an increase in transmissibility and a higher likelihood of escaping immune response after vaccination [4,7-12].\nIt is therefore important to have new surveillance metrics for monitoring the odds of having domestic cluster infection transmitted from few imported cases and setting up the alert threshold for forestalling community-acquired outbreaks, as traditional surveillance metrics, like effective reproductive number (Rt), are tailored for assessing the spread and control of community-acquired outbreaks at the population level, which may only involve a single country and a specific SARS-CoV-2 strain in a short period rather than the country of the imported case across the world and the full spectrum of SARS-CoV-2 strains with a long period [13-16]. Such a traditional epidemic surveillance model (eg, the SEIR [Susceptible-Exposed-Infected-Recovery] model) is not only limited to model the relationship of sparse cases of domestic cluster infection and small samples of imported cases from each original country, but also inflexible to make allowance for the heterogeneity of the imported-domestic transmission mode across countries and SARS-CoV-2 strains across time, as well as the variation across local regions in question. To consider these issues of heterogeneity, it is therefore necessary to develop new surveillance models and their corresponding metrics with a new statistical approach, such as a sampling method of machine learning, particularly the Bayesian Markov Chain Monte Carlo (MCMC) method, in conjunction with a sparse event history regression model, such as the extra-Poisson (random-effect) regression model with relevant parameters and random variables parameterized under the directed acyclic graphic (DAG) diagram.\nDeveloping these new surveillance metrics for quantifying the effect size of the transmission from importation to domestic cluster infection would not only be helpful for alerting emerging community-acquired outbreaks, but also aid health professionals having rapid preparedness of SARS-CoV-2 strain–dependent containment measures, including effective and efficient contact tracing. Using a series of chronological epidemic data on COVID-19 divided into 2 phases (non-VOC phase [wild type and D614G] and VOC phase) in Taiwan, this study aimed to develop new surveillance metrices across the periods of various SARS-CoV-2 strains for alerting emerging community-acquired outbreaks by monitoring the risk of small domestic cluster infections originating from the transmission of few imported cases of emerging variants in order to forestall community-acquired outbreaks when facing emerging SARS-CoV-2 variants. The Bayesian MCMC sampling method was therefore used to estimate and predict the new surveillance metrics underpinning the DAG diagram of the Poisson or negative binomial random-effect regression model.","[SUBTITLE] Data Sources [SUBSECTION] Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\nPublicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\n[SUBTITLE] Containment Measures for the Non-VOC Phase in Taiwan [SUBSECTION] The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\nThe containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\n[SUBTITLE] Containment Measures for the VOC Phase in Taiwan [SUBSECTION] The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\nThe Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\n[SUBTITLE] Statistical Analysis [SUBSECTION] [SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\n[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).","Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].","The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.","The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.","[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).","We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).","The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\nis proportional to the product of the kernel distribution written by\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\n3. Compute the acceptance probability\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).","[SUBTITLE] Evaluation of the Optimal Time Lag Model [SUBSECTION] After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\nAfter the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\n[SUBTITLE] Surveillance Metrics for the Imported-Domestic Transmission Mode [SUBSECTION] The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nThe weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\n[SUBTITLE] External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand [SUBSECTION] To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\nTo validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.","After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).","The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\naVOC: variant of concern.\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.","During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.","There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.","After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.","To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.","Many cyclical community-acquired outbreaks in each country or region during the COVID-19 pandemic have been noted from 2020 to 2022, and these epidemics have occurred in parallel with the evolution of various emerging SARS-CoV-2 variants, including the wild-type/D614G strain during the non-VOC phase and the Alpha, Beta, Gamma, Delta, and Omicron strains during the VOC phase. More importantly, when facing emerging variants, there were corresponding chains of containment measures with the following 3 serial steps: (1) border control of imported cases together with quarantine and isolation; (2) contact tracing and epidemic investigation of domestic cluster infection together with testing for detecting the foci of infection earlier (small households to large institutions); and (3) control of large-scale community-acquired infection with population-based approaches, mainly involving NPIs and mass vaccination. Although most countries focus on steps 1 and 2 for averting community-acquired outbreaks in the beginning, they end up having no choice but to adopt step 3 involving population-based approaches. Accordingly, most epidemic surveillance models still follow traditional surveillance metrics like the effective basic reproductive number (Rt) for containing community-acquired outbreaks. However, it is still important to develop a new surveillance model with new surveillance metrics commensurate with steps 1 and 2 for forestalling community-acquired outbreaks when facing emerging SARS-CoV-2 variants like the updated Omicron subvariants BA.4/BA.5. Most importantly, such a new surveillance model with useful metrics can be robustly applied across countries and time, covering various emerging SARS-CoV-2 variants detected from imported cases.\nUsing Taiwan empirical data, the proposed new surveillance model for monitoring cluster infections in the wake of imported cases was assessed in 5 periods (wild-type, D614G, Alpha VOC, Delta VOC, and Omicron VOC). The first metric of Dci/Imc per week was used to estimate the effect size of the risk of infection through imported-domestic transmission. The second metric involving the upper bound of the 95% CrI for predicted domestic cases of cluster infections derived from imported cases 1 week before provided the alert threshold for guiding the preparedness of containment measures for preventing community-acquired outbreaks in each country or region. Such an alert threshold would be affected by the characteristics of each emerging SARS-CoV-2 variant, as well as the underlying coverage rate of vaccination and the extent of NPIs. By using empirical data on imported and domestic COVID-19 cases in Taiwan, the effect size of Dci/Imc and the alert threshold were estimated as 9.54% and 12.59% for the wild-type/D614G strain, 14.14% and 25.10% for the Alpha VOC, 10.05% and 13.32% for the Delta VOC, 4.59% and 5.75% for the Omicron BA.1 VOC, and 4.29% and 5.19% for the Omicron BA.2 VOC, respectively, in 2 periods. It should be noted that the interpretation of the absolute effect size of Dci/Imc across various emerging SARS-CoV-2 variants should be taken with great caution.\nWhen a similar logic is applied to the alert threshold, the threshold value for weekly domestic cases during 3 SARS-CoV-2 variant periods would not go beyond 20 cases under the low coverage rate of vaccination and good performance of NPIs before the Omicron VOC period. After the Omicron VOC period, the alert threshold for weekly domestic cases would be 180 cases under the high vaccination rate and minimal NPIs. Empirical evidence on whether and how community-acquired outbreaks can be averted through different periods of SARS-CoV-2 variants with the proposed new surveillance model has been demonstrated by Taiwan data. Outbreaks were averted during the wild-type and D614G periods. In contrast, large-scale outbreaks could not be averted during the Alpha VOC period when the expected number of domestic cases was far beyond the alert threshold for an outbreak between May 9 and May 12, 2021, because of the increased transmissibility of the Alpha VOC that was supported by the increased risk of imported-domestic transmission in comparison with the wild-type/D614G variant (14.14% vs 9.54%). However, a low level of NPIs might also have contributed to such an outbreak around mid-May 2021 (30%; Figure 5). In the Delta VOC period after excluding the outbreak related to the Alpha VOC in Taiwan, the level of the NPI alert and the strict border control strategies implemented since the outbreak period reduced the risk of imported-domestic transmission to 10.05% (Table 1) and averted a community-acquired outbreak of the Delta VOC.\nGiven the high coverage of full vaccination, lower estimates of Dci/Imc for the Omicron VOC were seen, ranging from 4.3% to 4.6%. As the protective effect of the Oxford/AstraZeneca vaccine in particular started to wane in the community, the observed number of domestic cases was beyond the threshold of outbreak during the early Omicron BA.1 VOC period. Guided by the alert threshold, several containment strategies, including more restricted border control and rapid RT-PCR testing on arrival for travelers, and rapid booster shots for eligible adults, were implemented to avert a large-scale outbreak. However, another large-scale community-acquired outbreak could not be averted in late March 2022 owing to the observed cases going beyond the alert threshold partially due to waning of the protective effect of the mRNA-1273 vaccine (Moderna) or BNT162b2 vaccine (Pfizer–BioNTech).\nAlthough our metrics of the risk for domestic cluster infection and the alert threshold are pivotal in imported cases 1 week before applying the surveillance model for monitoring imported cases 1 week prior to the formation of cluster infections of domestic COVID-19 cases, they are very useful for alerting the surrounding community in proximity to imported cases beyond the threshold of the upper bound of the 95% CrI to enhance NPIs and active rapid testing with effective contact tracing and epidemic investigation for the observed cases. This accounted for the lack of community-acquired outbreaks before mid-May 2021 in Taiwan. Such good control over the COVID-19 epidemic has been reported in previous studies by evaluating NPIs at the individual and population levels [5,19,26], and using the traditional surveillance model for assessing the duration from Rt larger than 1 to Rt smaller than 1 and the case load following the machining learning model [27].\nSeveral previous studies have proposed an early warning model in relation to contact tracing and epidemic investigation before community-acquired outbreaks. However, our study differs from 2 recent previous studies [15,16] that developed an early warning model of COVID-19 outbreaks, in 2 main aspects. First, both studies covered a short period that reflected 1 or 2 SARS-CoV-2 strains and used data based on a single country. They were therefore unable to test the robustness of their models for a series of SARS-CoV-2 variants and samples across countries. Guan et al used human mobility data in Israel over the period from February 1, 2020, to January 7, 2021. Specifically, they trained their model over the period from April 6 to October 24, 2020, and evaluated the model’s predictive ability over 2 very short periods (November 1-30, 2020, and December 1-31, 2021) [15]. Kogan et al employed data in the United States that had been obtained from multiple digital traces over 2 short periods (March 1-May 31, 2020, and June 1-September 30, 2020) [16]. In contrast, the proposed new surveillance model made use of the full chronological empirical data in Taiwan from January 1, 2020, to April 2, 2022, covering various emerging SARS-CoV-2 variants. Our model is robust across a long period involving various variants and across countries covering different geographical and cultural conditions when using imported cases. Second, the data derived from digital traces, for example, Google Trends, used in the studies by Guan et al and Kogan et al may be affected by media activities. Furthermore, it is not easy to obtain accurate real-time data in countries with unavailable technological infrastructure, strong information censoring, and a lack of transparency. Instead of focusing on digital traces, we made use of imported cases that may be less likely to be affected by confounding factors. The use of imported data in our proposed surveillance model can be applied to countries with different political and social conditions and at different technological development stages. Moreover, our study is highly relevant to health regulators and public health policy makers, particularly in countries that have opened their borders and eased/removed NPI measures.\nIn addition to the illustration of the Taiwan experience, the external validation involving New Zealand further adds credibility to the application of this surveillance model in a scenario without an outbreak. This model can also be applied to those countries with controllable community-acquired outbreaks, such as Israel [28] and Qatar [29], after the mass vaccination program since early 2021, to monitor the impact of imported cases on the risk of domestic cluster infection. This is especially important for outbreaks resulting from vaccine breakthrough in countries or regions with high vaccine coverage, such as Singapore [30] and Israel [31], or the rapid waning of booster effectiveness worldwide, possibly affecting the community-level spread of SARS-CoV-2 VOCs [32].\nOne of the major limitations of the current model pertains to the generalization of the proposed new surveillance model. There are 2 major circumstances that require the refinement of the current proposed epidemic surveillance model. The proposed model has not incorporated health care capacity for accommodating the threshold of tolerable COVID-19 cases responsible for each episode of the outbreak. In consideration of resuming prepandemic activity, making allowance for this factor is of paramount importance for the implementation of NPIs and testing given the vaccine coverage rate. Different countries and regions may require different outbreak thresholds based on this global surveillance model. With increasing cases of vaccine breakthrough; the rapid emergence of VOCs with a wide spectrum of immunogenicity, high transmissibility, and resistance to antibodies associated with natural infection or vaccination; and the waning of immunity in the population, there is a high likelihood for the continued spread of SARS-CoV-2 in the population [33]. Given the possibility of a long-term association between SARS-CoV-2 and the human population, the goal of epidemic surveillance may shift from the elimination of this pathogen to a balance among health care capacity, socioeconomic activity, and population immunity. If this occurs, the proposed surveillance model should take this factor into account and should be used as a guide to inform the containment measures required to mitigate large-scale outbreaks according to health care capacity. Moreover, as the border control policy on quarantine and isolation of imported cases gets altered with the advent of high-performance rapid testing and the gradual expansion of vaccine coverage worldwide, the surveillance model for monitoring imported-domestic transmission to avert outbreaks may vary from country to country, depending on the extent of NPIs, administration of tests, coverage rate of vaccines, and administration of vaccine boosters. Such a heterogeneity should be taken into account to refine the surveillance model on imported-domestic transmission when it is applied to avert a large-scale outbreak. More importantly, our new surveillance model and metrics are not meant to replace conventional surveillance and corresponding metrics like Rt for assessing how to eliminate the spread of large-scale community-acquired outbreaks. When a community-acquired outbreak occurs, the conventional surveillance SEIR model is needed to assess the effectiveness of containment measures, as shown in Figure 5. Based on the SEIR model, the Rt decreased from 4.0 to 0.7 from May 18, 2021, to July 31, 2021. The effectiveness of NPIs and testing, which reflects the strategies implemented 2 weeks ago, was over 60% after May 26, 2021, and increased to over 90% after June 14, 2021. A similar finding was noted for a community-acquired outbreak of Omicron BA.2. Again, Rt reduced from 7.7 (value of Rt in early January) to less than 1 (around the end of January; Figure 5B). Our proposed new surveillance model has a supplementary role as a global vigilance method for forestalling large-scale local community-acquired outbreaks of emerging SARS-CoV-2 VOCs in each country and region worldwide.\nIn conclusion, a global new surveillance model and metrics have been proposed for monitoring imported cases of SARS-CoV-2 variants from the non-VOC phase to the VOC phase, using the Taiwan scenario. The new surveillance model and metrics are very useful for forestalling a new large-scale community-acquired outbreak through monitoring of the imported-domestic transmission mode associated with emerging infectious diseases in the future."],"string":"[\n \"During the COVID-19 pandemic lasting for over 2.5 years, countries around the world have experienced cyclical COVID-19 changes alternating between lifting and operating nonpharmaceutical interventions (NPIs) and between the protective and waning effects of vaccines when facing the incessant epidemics of the COVID-19 pandemic [1-4]. The cyclical resurgence of COVID-19 at the country, continental, and global levels is mainly caused by emerging SARS-CoV-2 variants, particularly variants of concern (VOCs). In response to the resurgence of community-acquired outbreaks, 2 containment measures have become important, including the timely adjustment of NPIs (strengthened border control strategies and restricted social activities) combined with testing and the launch of mass primary and booster vaccinations [5-7].\\nIt should be noted that the typical pattern of transmission from an imported case to domestic cluster infection is often the root of local community-acquired outbreaks caused by emerging SARS-CoV-2 variants from any region or country across the globe [2,7-9]. Such an importation-cluster transmission mode has been clearly demonstrated by the resurgence of global epidemic waves following the emergence of dominant strains of Alpha, Beta, Gamma, Delta, and Omicron VOCs. To avert local community-acquired outbreaks of emerging SARS-CoV-2 variants, rapid preparedness of containment measures and effective contact tracing are mandatory when domestic cluster infections are identified after the introduction of emerging imported cases. In addition, the risk of domestic cluster infection on the introduction of imported cases varies with each emerging SARS-CoV-2 strain owing to the evolutionary characteristics of invading VOCs, including an increase in transmissibility and a higher likelihood of escaping immune response after vaccination [4,7-12].\\nIt is therefore important to have new surveillance metrics for monitoring the odds of having domestic cluster infection transmitted from few imported cases and setting up the alert threshold for forestalling community-acquired outbreaks, as traditional surveillance metrics, like effective reproductive number (Rt), are tailored for assessing the spread and control of community-acquired outbreaks at the population level, which may only involve a single country and a specific SARS-CoV-2 strain in a short period rather than the country of the imported case across the world and the full spectrum of SARS-CoV-2 strains with a long period [13-16]. Such a traditional epidemic surveillance model (eg, the SEIR [Susceptible-Exposed-Infected-Recovery] model) is not only limited to model the relationship of sparse cases of domestic cluster infection and small samples of imported cases from each original country, but also inflexible to make allowance for the heterogeneity of the imported-domestic transmission mode across countries and SARS-CoV-2 strains across time, as well as the variation across local regions in question. To consider these issues of heterogeneity, it is therefore necessary to develop new surveillance models and their corresponding metrics with a new statistical approach, such as a sampling method of machine learning, particularly the Bayesian Markov Chain Monte Carlo (MCMC) method, in conjunction with a sparse event history regression model, such as the extra-Poisson (random-effect) regression model with relevant parameters and random variables parameterized under the directed acyclic graphic (DAG) diagram.\\nDeveloping these new surveillance metrics for quantifying the effect size of the transmission from importation to domestic cluster infection would not only be helpful for alerting emerging community-acquired outbreaks, but also aid health professionals having rapid preparedness of SARS-CoV-2 strain–dependent containment measures, including effective and efficient contact tracing. Using a series of chronological epidemic data on COVID-19 divided into 2 phases (non-VOC phase [wild type and D614G] and VOC phase) in Taiwan, this study aimed to develop new surveillance metrices across the periods of various SARS-CoV-2 strains for alerting emerging community-acquired outbreaks by monitoring the risk of small domestic cluster infections originating from the transmission of few imported cases of emerging variants in order to forestall community-acquired outbreaks when facing emerging SARS-CoV-2 variants. The Bayesian MCMC sampling method was therefore used to estimate and predict the new surveillance metrics underpinning the DAG diagram of the Poisson or negative binomial random-effect regression model.\",\n \"[SUBTITLE] Data Sources [SUBSECTION] Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\\nPublicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\\n[SUBTITLE] Containment Measures for the Non-VOC Phase in Taiwan [SUBSECTION] The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\\nThe containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\\n[SUBTITLE] Containment Measures for the VOC Phase in Taiwan [SUBSECTION] The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\\nThe Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\\n[SUBTITLE] Statistical Analysis [SUBSECTION] [SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\\n[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\",\n \"Publicly available information on COVID-19, including the daily number of cases, recovered patients, and deaths from January 1, 2020, to April 2, 2022, in Taiwan, was extracted from the report of the Central Epidemic Command Centre and the Taiwan National Infectious Disease Statistics System maintained by the Taiwan Centre for Disease Control [17]. During the period between January 11 and June 20, 2020, tabular data with epidemiologic information on COVID-19 mentioned above by county and origin of cases (domestic versus imported) were obtained. After October 2020, only aggregated numbers of imported and domestic COVID-19 cases without detailed information at the county and city levels were provided. The population sizes of 23 counties and cities in Taiwan were extracted from the official website of the Department of Household Registration [18].\",\n \"The containment measures for the non-VOC phase in Taiwan centered on 2 strategies, namely border control with quarantine and isolation, and NPIs without various strategies. Multimedia Appendix 1 shows the timelines of the evolution of border control measures for this non-VOC phase. The number of imported and domestic cases of COVID-19 by the date of onset on a weekly basis is presented in Figure 1.\\nCOVID-19 epidemics in Taiwan for periods without outbreaks by the origin of cases (imported vs domestic).\\nThe COVID-19 cases in Taiwan in 2020 mainly included imported cases (Figure 1). The risk of an outbreak following the transmission of COVID-19 to the community from these imported cases was largely reduced by the very strict border control strategies with quarantine and isolation in conjunction with NPIs, such as wearing masks and social distancing [19]. Multimedia Appendix 2 provides details on the criteria and guidelines for the implementation of 4 COVID-19 alert levels to target outbreaks in Taiwan.\",\n \"The Alpha VOC became the predominant strain of the global pandemic by the end of 2020. Several cluster infections occurred in hospitals and households since January 2021, but were still under control until mid-May 2021, when a large-scale outbreak of the Alpha VOC occurred. On the top of border control measures with quarantine and isolation implemented since the non-VOC phase in Taiwan, the focus of containment measures for averting community-acquired outbreaks turned to community-based active surveillance with rapid test stations for the hotspots of outbreaks and enhanced NPIs, including strict regulation for wearing masks, restriction of public gathering, setting up of check points for high-risk areas such as public transportation sites and markets, and restriction of nonessential services such as restaurants and pubs. Multimedia Appendix 1 summarizes the timeline of the implementation of a series of containment measures for the VOC phase starting from the enhancement of NPIs from level 1 to level 2 alert until high restriction border control for travelers. The level 3 alert was rapidly extended to a nationwide level 3 alert on May 19, 2021 [20,21]. During the Delta VOC period (from August to December 2021) and Omicron VOC period (December 2021 onward), transmission in the community has been threatened by imported cases. In addition to containment measures, high coverage of vaccination has been an effective prevention strategy during these 2 periods. In response to the rapid spread of the Omicron VOC, inbound passengers have to follow updated regulations with more frequent reverse transcription-polymerase chain reaction (RT-PCR) testing plus rapid testing, and a possible mandatory 14-day quarantine based on the vaccination status. Additionally, inbound passengers have to provide negative COVID-19 RT-PCR test reports within 2 days and have to take a government-funded rapid RT-PCR test on arrival starting January 11, 2022. Owing to waning of the effects of vaccines, booster shots have been allowed for all adults who have received 2 vaccine doses for 12 weeks (84 days), since January 7, 2022.\",\n \"[SUBTITLE] New Surveillance Metrics for Quantifying Imported-Domestic Transmission [SUBSECTION] We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\nWe used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\\n[SUBTITLE] Estimation With the Bayesian MCMC Method [SUBSECTION] The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\\nThe Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\",\n \"We used an extra-Poisson regression model with a Bayesian DAG approach [22] to calculate the expected weekly domestic cluster infections associated with imported cases of COVID-19, as shown in the right panel of Multimedia Appendix 3. For the jth county or city with the Ytj domestic case at week t, the extra-Poisson regression model can be specified by\\nwhere offestj is the population of log scale, and the heterogeneity of imported-domestic transmission across counties and cities in Taiwan is captured by a normal distributed random intercept parameter, αj. While the common intercept parameter, α0, represents the common risk of transmission in Taiwan, the heterogeneity is captured by the variance parameter, σ2α. With this framework, the number of cases arising from domestic cluster infection caused by imported cases per week before (Xj,t-1) can be assessed by using the regression coefficient β, which becomes the first surveillance metric and is denoted as Dci/Imc per week for estimating the effect size of domestic cluster infection. The larger the value of this metric estimated, the larger the domestic cluster infection. The extra variation across cities and counties regarding the transmission of COVID-19 associated with imported cases was captured by a random effect (αj) incorporated into the Poisson regression model, which is also called the random-effect Poisson regression model. The predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 3) can be generated by using the number of imported cases in the current week (X[t]; Multimedia Appendix 3) in conjunction with the posterior distribution of the force of transmission (β), standing for the metric of Dci/Imc per week, and the common intercept (α0) taking into account the county-level heterogeneity of COVID-19 transmission (σ2α). The Poisson model has been widely applied to sparse counts of domestic infection, which occur independently if there is a lack of larger cluster infections, with a high potential of developing into a large-scale community-acquired outbreak. If the observed value of our model is beyond the upper limit of the 95% credible interval (CrI), it means that sparse and independent assumptions based on the Poisson distribution are violated and implies a high potential of yielding a large-scale community-acquired outbreak. Accordingly, the second surveillance metric is to build up the alert threshold of emerging community-acquired outbreaks and to provide guidance for the rapid preparedness of containment measures (including effective and efficient contact tracing) for forestalling community-acquired outbreaks.\\nAs mentioned above, data were divided into the non-VOC phase and VOC phase. The former period used for estimating the parameters of the following extra-Poisson regression model was based on imported and domestic cases between January 11 and June 20, 2020, covering the wild-type and D614G period in Taiwan. Because imported cases require an incubation time to generate secondary cases, we tested the lag time of imported cases by 0 weeks (concurrent, Xjt), 1 week (Xj,t-1), and 2 weeks (Xj,t-2), and further selected the optimal lag time interval with the smallest deviance information criterion (DIC).\\nRegarding the impact of imported cases on the occurrence of domestic cases for the early Alpha (October 11, 2020, to May 12, 2021), Delta (August 8 to December 9, 2021), and Omicron (December 12, 2021, to April 2, 2022) VOC periods without outbreaks in Taiwan, a Bayesian negative binomial regression model was applied to take into account the heterogeneity across counties and cities associated with the imported-domestic transmission of COVID-19 owing to the lack of detailed information on the cases in counties and cities. Multimedia Appendix 4 shows the DAG model for assessing the force of imported-domestic transmission by using a Bayesian negative binomial regression model. Following the approach applied for the wild-type and D614G period, the 1-week lag model was adopted. For week t, the number of cases Yt resulting from imported cases 1 week prior, Xt-1, can be modeled by using the negative binomial regression model as follows:\\nwhere the heterogeneity is captured by the dispersion parameter 1/k. Similar to the extra-Poisson regression model as above, the risk of imported-domestic transmission can thus be assessed by using the regression coefficient β for estimating the effect size of Dci/Imc. Following the extra-Poisson approach, the predicted distribution of the number of expected domestic cases in the next week (μ.pred[t+1]; Multimedia Appendix 4) for the Bayesian negative binomial model can be generated from the current number of imported cases (X[t]; Multimedia Appendix 4) by using the posterior distribution of imported-domestic transmission (β) and the dispersion parameter (1/k).\",\n \"The Bayesian MCMC method was used to generate the samples derived from the posterior distributions of parameters for estimating 2 surveillance metrics. With the Markov chain underpinning, a stationary distribution for parameters can be reached in the long run under regular conditions. Independent samples can thus be generated from such a stationary posterior distribution on the basis of which inferences can be made [23]. The DAG models depicted in Multimedia Appendix 3 and Multimedia Appendix 4 were applied to facilitate the decomposition of joint distribution into full conditional density distribution by using the relationship between parent and child nodes [24]. Taking the extra-Poisson regression model as an example, the joint distribution,\\nis proportional to the product of the kernel distribution written by\\nIn our application, noninformative priors were used to derive the samples from the stationary posterior distribution of parameters, including the risk of imported-domestic transmission (β), the common intercept (α0), and the county-specific random effect (σα).\\nA block-wise Metropolic-Hasting sampler was applied to generate samples from the stationary posterior distribution. The sampling algorithm is detailed as follows:\\n1. Start with an initial value (β(0), α(0), α0(0), σ2α(0)) selected from the support of each parameter.\\n2. Draw the candidate value for the first parameter, say β(1), from a normal proposal distribution, q(β).\\n3. Compute the acceptance probability\\n4. Draw u from uniform (0,1) and update β(0) with β(1) if u < r(β(1), β(0) | α(0), α0(0), σ2α(0)); otherwise, repeat steps 2 and 3.\\n5. Draw the candidate value for the next parameter, α(1), to update the parameter sample with (α(0) | α0(0), σ2α(0), β(1)) by using steps 2 to 4.\\n6. Repeat steps 2 to 5 for the rest of the parameters, (α0, σ2α), to derive (β(1), α(1), α0(1), σ2α(1)) to complete an iteration of the update for parameter samples.\\nThinning intervals of 10 and 100,000 iterations were used to generate the 10,000 posterior samples after 250,000 burn-in iterations by using the Bayesian MCMC methods mentioned above.\\nWe estimated the effect size of Dci/Imc per week for each period corresponding to the type of SARS-CoV-2 variant. We built up the alert threshold by using the upper limit of the 95% CrI of the predicted number of domestic cases (μ.pred[t+1]; Multimedia Appendix 3 and Multimedia Appendix 4) generated by the parameters after updating the data on the non-VOC phase in Taiwan for alerting the possibility of yielding a large-scale community-acquired outbreak through imported-domestic transmission in the subsequent epochs. The possibility of a community-acquired outbreak was deemed low if observed domestic cases were not more than the alert threshold, namely the upper limit of the 95% CrI. Otherwise, an outbreak was likely to occur, and therefore, the rapid preparedness of containment measures, including effective and efficient contact tracing, would be flagged to forestall the ensuing community-acquired outbreak.\\nTo validate the proposed surveillance model for the transmission from imported to domestic cases during the non-VOC period in Taiwan, the publicly available COVID-19 data provided by the Ministry of Health in New Zealand were used [25]. The chronological order of the incidence of COVID-19 for the hotspots was compared to validate the epidemic surveillance model for an outbreak.\\nAll statistical analyses were performed using SAS 9.4 software (SAS Institute Inc).\",\n \"[SUBTITLE] Evaluation of the Optimal Time Lag Model [SUBSECTION] After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\\nAfter the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\\n[SUBTITLE] Surveillance Metrics for the Imported-Domestic Transmission Mode [SUBSECTION] The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\\naVOC: variant of concern.\\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\nThe weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\\naVOC: variant of concern.\\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\n[SUBTITLE] External Validation of the Surveillance Metrics for Domestic Cluster Infections Using Imported Cases in New Zealand [SUBSECTION] To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\\nTo validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\",\n \"After the application of the Bayesian MCMC method for the identification of the optimal time (in weeks) lag of imported cases prior to cases arising from domestic cluster infection, the 1-week lag of imported cases yielded a DIC of 255.8, which was smaller than the DICs of the model with concurrent week imported cases (260.3) and the model with a 2-week lag of imported cases (279.5).\",\n \"The weekly observed number (red dot) and expected number (green circle) of domestic cases are shown in Figure 2 (wild-type and D614G period, January to September 2020), Figure 3 (Alpha VOC period, October 2020 to May 2021), and Figure 4 (Delta VOC period, mid-August to mid-December 2021; and Omicron VOC period, mid-December 2021 to early-April 2022). Table 1 shows the details of the estimated results of the parameters encoded in the Bayesian extra-Poisson regression model with a 1-week lag of imported cases regarding the 3 periods without outbreaks in Taiwan, namely the wild-type and D614G period, early Alpha VOC period, and Delta VOC period.\\nThe upper bound of the 95% CrI of expected cases (dotted line, Figures 2-4) has been plotted to provide the alert threshold of domestic cluster infection in the community caused by transmission from imported cases 1 week before. This 1-week prior alert on the risk of elevated Dci/Imc per week guided the vigilance on NPIs for averting further community-acquired outbreaks.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the wild-type/D614G period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Alpha variant of concern period.\\nNumber of observed (dotted point) and expected (green circle) domestic cases with the upper limit of the 95% credible interval (CrI) (dotted line) by week in the Delta (August 8, 2021, to December 9, 2021) and Omicron (December 12, 2021, to April 2, 2022) variant of concern periods.\\nEstimated results for the risk of imported-domestic transmission of COVID-19 for 3 periods in Taiwan.\\naVOC: variant of concern.\\n[SUBTITLE] Wild-Type and D614G Period [SUBSECTION] During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\nDuring the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\\n[SUBTITLE] Alpha VOC Period [SUBSECTION] There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\nThere had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\\n[SUBTITLE] Delta and Omicron VOC Period [SUBSECTION] After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\\nAfter controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\",\n \"During the wild-type/D614G period, the estimated Dci/Imc per week was 9.54% (95% CrI 6.44%-12.59%; Table 1). Figure 2 shows that there were 5 weeks (January 26 to February 1, February 2 to February 8, February 16 to February 22, February 23 to February 29, and March 15 to March 21, 2020) in which the observed numbers of domestic cases exceeded the alert thresholds. This period yielded 81% of clustered cases (22 of 27 community-acquired cases) in 5 clusters in Taiwan, including 3 household clusters (with 5, 3, and 6 COVID-19 cases, respectively), 1 medical institute cluster (with 9 COVID-19 cases), and 1 academic institute cluster (with 4 COVID-19 cases). Guidance of the alert thresholds from this early period of the wild-type COVID-19 strain provided a strong rationale for being on alert for the ensuing cluster infections from the preceding 1 week when imported cases were introduced. This accounted for why none of these 5 cluster events led to any large-scale community-acquired outbreaks in Taiwan. There was rapid preparedness of containment measures with strict NPIs together with effective and efficient contact tracing of all possible susceptible individuals.\\nFigure 2 shows that this surveillance metric was very useful, particularly when there was a substantial surge in imported cases resulting from the large-scale COVID-19 pandemic worldwide. This could be seen in our cases between March and April 2020, as shown in Figure 1. Again, the surveillance metric was used for alerting about possible cluster infections to forestall further community-acquired outbreaks. Alerted by the threshold (20 domestic cases per week), the observed domestic cases were kept lower than the alert threshold to avoid large-scale outbreaks in April. Since then, there had not been any domestic case until December 2020.\",\n \"There had been no outbreak during the early Alpha VOC phase of the COVID-19 pandemic from October to December 2020. The second surge of imported cases occurred from January 2021 onwards (Figure 1). Again, there was 1 week (January 17 to January 23, 2021) in which the observed number of domestic cases was beyond the alert threshold, resulting from hospital-based cluster infections and 3 subsequent household clusters (11 family members) (Figure 3). The source of this cluster infection was later identified as an imported case infected with the Alpha VOC. After being alerted by the proposed surveillance metric and following timely contact tracing and containment measures, including quarantine and isolation for all staff members in the hospital and their close contacts for 14 days, there was no further outbreak until early May 2021, when the number of observed domestic cases reached beyond 20, which was higher than the expected surveillance curve (5 cases) and the corresponding upper bound of the 95% CrI (14 cases). The estimated results on the basis of the empirical data during this Alpha VOC period showed that the Dci/Imc increased to 14.14% (95% CrI 5.41%-25.10%; Table 1). The peak of domestic cases far beyond the threshold value in early May 2021 not only presaged the ensuing outbreaks, but also revealed the loose NPIs at that time without an available vaccine in Taiwan. If effective contact tracing and timely containment measures had been deployed in advance on the basis of the increased risk and the alert threshold in the week between May 2 and May 8, the ensuing community-acquired outbreak involving the Alpha VOC could have been prevented (bottom panel of Figure 1). This outbreak lasted for 2.5 months and subsided in July 2021. The corresponding periods of the estimated results for the effectiveness of NPIs and testing implemented in Taiwan during the Alpha VOC phase are shown in Figure 5A.\\nEffective reproductive number (Rt) and effectiveness of nonpharmaceutical interventions (NPIs) and testing in Taiwan. (A) Alpha variant. (B) Omicron variant.\",\n \"After controlling the outbreak in the Alpha VOC period, there was monitoring of domestic cluster infections for the Delta VOC since August 8, 2021, and the Omicron VOC from December 10, 2021. After mid-August, there had been a cluster infection during the Delta VOC phase of the COVID-19 pandemic. The estimated results (Table 1, Delta VOC period) showed that the Dci/Imc remained at 10.01% (95% CrI 6.85%-13.32%), although the Delta VOC had higher transmissibility and escaped vaccine-induced immunity. The left panel of Figure 4 shows that for 2 weeks (August 22 to August 28, 2021, and September 19 to September 25, 2021) when a surge in the number of domestic cases as a result of Dci/Imc was expected (green circle), the observed number of domestic cases was far below the threshold of outbreak. This can be attributed to the implementation of enhanced containment measures, including the strengthening of border control strategies with multiple tests on arrival and during quarantine, the collective quarantine strategy, and the elevated alerts of NPIs to levels 2 and 3 since the outbreak in May 2021 in Taiwan. Since then, the weekly observed cases were below the alert threshold owing to NPIs with a level 2 alert and the rapid administration of vaccines.\\nThe right panel of Figure 4 shows the observed and predicted numbers of cases along with the alert threshold during the Omicron VOC period. Given the increasing coverage rate of vaccination, the risk of domestic cluster infection per imported case for the Omicron BA.1 VOC reduced to 4.59% (95% CrI 3.66%-5.75%), but an upsurge in domestic cases was still observed because the Omicron BA.1 VOC was considered to have a high transmission probability. In the week from January 9 to January 15, 2022, the observed number exceeded the alert threshold, indicating a high potential for a community-acquired outbreak. There was indeed a small-scale community-acquired outbreak of the Omicron BA.1 VOC. After a series of containment measures, including rapid RT-PCR tests for inbound passengers on arrival coupled with stringent quarantine and isolation, rapid booster vaccination, and enhanced NPIs with a level 2 alert in the community, the community-acquired outbreak subsided by the end of February 2022 (Figure 5).\\nThere was a return to the imported-domestic transmission model, with the surveillance metric of Dci/Imc estimated at 4.29% (95% CrI 3.52%-5.19%) from February until March 20, 2022. After that, a similar circumstance beyond the alert threshold was noted for the invasion of the imported Omicron BA.2 VOC, and a community-acquired outbreak started from March 20 to 26, 2022, resulting in a large-scale community-acquired outbreak from early April until July 2022.\",\n \"To validate the proposed model and extend its application to different periods of SARS-CoV-2 variants, the proposed extra-Poisson regression model was applied to data on the New Zealand COVID-19 outbreak in 2020. Multimedia Appendix 5 shows the estimated results obtained. Notably, in New Zealand, the risk of Dci/Imc per week increased to 9.38% (95% CrI 8.88%-9.86%), which was close to the estimated results based on Taiwan data (9.54%, 95% CrI 6.44%-12.59%) in the same period. Details regarding the spatial temporal distribution of COVID-19 outbreaks by types of cases in New Zealand are provided in Multimedia Appendix 6. Multimedia Appendix 7 shows the predicted number of domestic cases by using the parameters trained from the empirical data of New Zealand (Multimedia Appendix 5). Similar to the application in Taiwan, the risk of an outbreak associated with imported cases could be assessed by comparing the observed cases (red dot in Multimedia Appendix 7) with the alert threshold (dotted line in Multimedia Appendix 7). The detailed interpretation of the results of this external validation is elaborated in Multimedia Appendix 8.\",\n \"Many cyclical community-acquired outbreaks in each country or region during the COVID-19 pandemic have been noted from 2020 to 2022, and these epidemics have occurred in parallel with the evolution of various emerging SARS-CoV-2 variants, including the wild-type/D614G strain during the non-VOC phase and the Alpha, Beta, Gamma, Delta, and Omicron strains during the VOC phase. More importantly, when facing emerging variants, there were corresponding chains of containment measures with the following 3 serial steps: (1) border control of imported cases together with quarantine and isolation; (2) contact tracing and epidemic investigation of domestic cluster infection together with testing for detecting the foci of infection earlier (small households to large institutions); and (3) control of large-scale community-acquired infection with population-based approaches, mainly involving NPIs and mass vaccination. Although most countries focus on steps 1 and 2 for averting community-acquired outbreaks in the beginning, they end up having no choice but to adopt step 3 involving population-based approaches. Accordingly, most epidemic surveillance models still follow traditional surveillance metrics like the effective basic reproductive number (Rt) for containing community-acquired outbreaks. However, it is still important to develop a new surveillance model with new surveillance metrics commensurate with steps 1 and 2 for forestalling community-acquired outbreaks when facing emerging SARS-CoV-2 variants like the updated Omicron subvariants BA.4/BA.5. Most importantly, such a new surveillance model with useful metrics can be robustly applied across countries and time, covering various emerging SARS-CoV-2 variants detected from imported cases.\\nUsing Taiwan empirical data, the proposed new surveillance model for monitoring cluster infections in the wake of imported cases was assessed in 5 periods (wild-type, D614G, Alpha VOC, Delta VOC, and Omicron VOC). The first metric of Dci/Imc per week was used to estimate the effect size of the risk of infection through imported-domestic transmission. The second metric involving the upper bound of the 95% CrI for predicted domestic cases of cluster infections derived from imported cases 1 week before provided the alert threshold for guiding the preparedness of containment measures for preventing community-acquired outbreaks in each country or region. Such an alert threshold would be affected by the characteristics of each emerging SARS-CoV-2 variant, as well as the underlying coverage rate of vaccination and the extent of NPIs. By using empirical data on imported and domestic COVID-19 cases in Taiwan, the effect size of Dci/Imc and the alert threshold were estimated as 9.54% and 12.59% for the wild-type/D614G strain, 14.14% and 25.10% for the Alpha VOC, 10.05% and 13.32% for the Delta VOC, 4.59% and 5.75% for the Omicron BA.1 VOC, and 4.29% and 5.19% for the Omicron BA.2 VOC, respectively, in 2 periods. It should be noted that the interpretation of the absolute effect size of Dci/Imc across various emerging SARS-CoV-2 variants should be taken with great caution.\\nWhen a similar logic is applied to the alert threshold, the threshold value for weekly domestic cases during 3 SARS-CoV-2 variant periods would not go beyond 20 cases under the low coverage rate of vaccination and good performance of NPIs before the Omicron VOC period. After the Omicron VOC period, the alert threshold for weekly domestic cases would be 180 cases under the high vaccination rate and minimal NPIs. Empirical evidence on whether and how community-acquired outbreaks can be averted through different periods of SARS-CoV-2 variants with the proposed new surveillance model has been demonstrated by Taiwan data. Outbreaks were averted during the wild-type and D614G periods. In contrast, large-scale outbreaks could not be averted during the Alpha VOC period when the expected number of domestic cases was far beyond the alert threshold for an outbreak between May 9 and May 12, 2021, because of the increased transmissibility of the Alpha VOC that was supported by the increased risk of imported-domestic transmission in comparison with the wild-type/D614G variant (14.14% vs 9.54%). However, a low level of NPIs might also have contributed to such an outbreak around mid-May 2021 (30%; Figure 5). In the Delta VOC period after excluding the outbreak related to the Alpha VOC in Taiwan, the level of the NPI alert and the strict border control strategies implemented since the outbreak period reduced the risk of imported-domestic transmission to 10.05% (Table 1) and averted a community-acquired outbreak of the Delta VOC.\\nGiven the high coverage of full vaccination, lower estimates of Dci/Imc for the Omicron VOC were seen, ranging from 4.3% to 4.6%. As the protective effect of the Oxford/AstraZeneca vaccine in particular started to wane in the community, the observed number of domestic cases was beyond the threshold of outbreak during the early Omicron BA.1 VOC period. Guided by the alert threshold, several containment strategies, including more restricted border control and rapid RT-PCR testing on arrival for travelers, and rapid booster shots for eligible adults, were implemented to avert a large-scale outbreak. However, another large-scale community-acquired outbreak could not be averted in late March 2022 owing to the observed cases going beyond the alert threshold partially due to waning of the protective effect of the mRNA-1273 vaccine (Moderna) or BNT162b2 vaccine (Pfizer–BioNTech).\\nAlthough our metrics of the risk for domestic cluster infection and the alert threshold are pivotal in imported cases 1 week before applying the surveillance model for monitoring imported cases 1 week prior to the formation of cluster infections of domestic COVID-19 cases, they are very useful for alerting the surrounding community in proximity to imported cases beyond the threshold of the upper bound of the 95% CrI to enhance NPIs and active rapid testing with effective contact tracing and epidemic investigation for the observed cases. This accounted for the lack of community-acquired outbreaks before mid-May 2021 in Taiwan. Such good control over the COVID-19 epidemic has been reported in previous studies by evaluating NPIs at the individual and population levels [5,19,26], and using the traditional surveillance model for assessing the duration from Rt larger than 1 to Rt smaller than 1 and the case load following the machining learning model [27].\\nSeveral previous studies have proposed an early warning model in relation to contact tracing and epidemic investigation before community-acquired outbreaks. However, our study differs from 2 recent previous studies [15,16] that developed an early warning model of COVID-19 outbreaks, in 2 main aspects. First, both studies covered a short period that reflected 1 or 2 SARS-CoV-2 strains and used data based on a single country. They were therefore unable to test the robustness of their models for a series of SARS-CoV-2 variants and samples across countries. Guan et al used human mobility data in Israel over the period from February 1, 2020, to January 7, 2021. Specifically, they trained their model over the period from April 6 to October 24, 2020, and evaluated the model’s predictive ability over 2 very short periods (November 1-30, 2020, and December 1-31, 2021) [15]. Kogan et al employed data in the United States that had been obtained from multiple digital traces over 2 short periods (March 1-May 31, 2020, and June 1-September 30, 2020) [16]. In contrast, the proposed new surveillance model made use of the full chronological empirical data in Taiwan from January 1, 2020, to April 2, 2022, covering various emerging SARS-CoV-2 variants. Our model is robust across a long period involving various variants and across countries covering different geographical and cultural conditions when using imported cases. Second, the data derived from digital traces, for example, Google Trends, used in the studies by Guan et al and Kogan et al may be affected by media activities. Furthermore, it is not easy to obtain accurate real-time data in countries with unavailable technological infrastructure, strong information censoring, and a lack of transparency. Instead of focusing on digital traces, we made use of imported cases that may be less likely to be affected by confounding factors. The use of imported data in our proposed surveillance model can be applied to countries with different political and social conditions and at different technological development stages. Moreover, our study is highly relevant to health regulators and public health policy makers, particularly in countries that have opened their borders and eased/removed NPI measures.\\nIn addition to the illustration of the Taiwan experience, the external validation involving New Zealand further adds credibility to the application of this surveillance model in a scenario without an outbreak. This model can also be applied to those countries with controllable community-acquired outbreaks, such as Israel [28] and Qatar [29], after the mass vaccination program since early 2021, to monitor the impact of imported cases on the risk of domestic cluster infection. This is especially important for outbreaks resulting from vaccine breakthrough in countries or regions with high vaccine coverage, such as Singapore [30] and Israel [31], or the rapid waning of booster effectiveness worldwide, possibly affecting the community-level spread of SARS-CoV-2 VOCs [32].\\nOne of the major limitations of the current model pertains to the generalization of the proposed new surveillance model. There are 2 major circumstances that require the refinement of the current proposed epidemic surveillance model. The proposed model has not incorporated health care capacity for accommodating the threshold of tolerable COVID-19 cases responsible for each episode of the outbreak. In consideration of resuming prepandemic activity, making allowance for this factor is of paramount importance for the implementation of NPIs and testing given the vaccine coverage rate. Different countries and regions may require different outbreak thresholds based on this global surveillance model. With increasing cases of vaccine breakthrough; the rapid emergence of VOCs with a wide spectrum of immunogenicity, high transmissibility, and resistance to antibodies associated with natural infection or vaccination; and the waning of immunity in the population, there is a high likelihood for the continued spread of SARS-CoV-2 in the population [33]. Given the possibility of a long-term association between SARS-CoV-2 and the human population, the goal of epidemic surveillance may shift from the elimination of this pathogen to a balance among health care capacity, socioeconomic activity, and population immunity. If this occurs, the proposed surveillance model should take this factor into account and should be used as a guide to inform the containment measures required to mitigate large-scale outbreaks according to health care capacity. Moreover, as the border control policy on quarantine and isolation of imported cases gets altered with the advent of high-performance rapid testing and the gradual expansion of vaccine coverage worldwide, the surveillance model for monitoring imported-domestic transmission to avert outbreaks may vary from country to country, depending on the extent of NPIs, administration of tests, coverage rate of vaccines, and administration of vaccine boosters. Such a heterogeneity should be taken into account to refine the surveillance model on imported-domestic transmission when it is applied to avert a large-scale outbreak. More importantly, our new surveillance model and metrics are not meant to replace conventional surveillance and corresponding metrics like Rt for assessing how to eliminate the spread of large-scale community-acquired outbreaks. When a community-acquired outbreak occurs, the conventional surveillance SEIR model is needed to assess the effectiveness of containment measures, as shown in Figure 5. Based on the SEIR model, the Rt decreased from 4.0 to 0.7 from May 18, 2021, to July 31, 2021. The effectiveness of NPIs and testing, which reflects the strategies implemented 2 weeks ago, was over 60% after May 26, 2021, and increased to over 90% after June 14, 2021. A similar finding was noted for a community-acquired outbreak of Omicron BA.2. Again, Rt reduced from 7.7 (value of Rt in early January) to less than 1 (around the end of January; Figure 5B). Our proposed new surveillance model has a supplementary role as a global vigilance method for forestalling large-scale local community-acquired outbreaks of emerging SARS-CoV-2 VOCs in each country and region worldwide.\\nIn conclusion, a global new surveillance model and metrics have been proposed for monitoring imported cases of SARS-CoV-2 variants from the non-VOC phase to the VOC phase, using the Taiwan scenario. The new surveillance model and metrics are very useful for forestalling a new large-scale community-acquired outbreak through monitoring of the imported-domestic transmission mode associated with emerging infectious diseases in the future.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction","methods",null,null,null,null,null,null,"results",null,null,null,null,null,null,"discussion"],"string":"[\n \"introduction\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\"\n]"},"keywords":{"kind":"list like","value":["COVID-19","imported case","surveillance metric","early detection","community-acquired outbreak"],"string":"[\n \"COVID-19\",\n \"imported case\",\n \"surveillance metric\",\n \"early detection\",\n \"community-acquired outbreak\"\n]"}}},{"rowIdx":367,"cells":{"title":{"kind":"string","value":"The Risk of Hospitalization and Mortality After Breakthrough SARS-CoV-2 Infection by Vaccine Type: Observational Study of Medical Claims Data."},"pmid":{"kind":"string","value":"36265135"},"background_abstract":{"kind":"string","value":"Several risk factors have been identified for severe COVID-19 disease by the scientific community. In this paper, we focus on understanding the risks for severe COVID-19 infections after vaccination (ie, in breakthrough SARS-CoV-2 infections). Studying these risks by vaccine type, age, sex, comorbidities, and any prior SARS-CoV-2 infection is important to policy makers planning further vaccination efforts."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We performed Cox regression analysis to calculate the risk factors for developing a severe breakthrough SARS-CoV-2 infection in the study cohort by controlling for vaccine type, age, sex, comorbidities, and a prior SARS-CoV-2 infection."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"We found lower hazard ratios for those receiving the Moderna vaccine (P<.001) and Pfizer vaccine (P<.001), with the lowest hazard rates being for Moderna, as compared to those who received the Janssen vaccine, independent of age, sex, comorbidities, vaccine type, and prior SARS-CoV-2 infection. Further, individuals who had a SARS-CoV-2 infection prior to vaccination had some increased protection over and above the protection already provided by the vaccines, from hospitalization (P=.001) and death (P=.04), independent of age, sex, comorbidities, and vaccine type. We found that the top statistically significant risk factors for severe breakthrough SARS-CoV-2 infections were age of >50, male gender, moderate and severe renal failure, severe liver disease, leukemia, chronic lung disease, coagulopathy, and alcohol abuse."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Among individuals who were fully vaccinated, the risk of severe breakthrough SARS-CoV-2 infection was lower for recipients of the Moderna or Pfizer vaccines and higher for recipients of the Janssen vaccine. These results from our analysis at a population level will be helpful to public health policy makers. Our result on the influence of a previous SARS-CoV-2 infection necessitates further research into the impact of multiple exposures on the risk of developing severe COVID-19."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Male","COVID-19","Viral Vaccines","SARS-CoV-2","Vaccination","Hospitalization"],"string":"[\n \"Humans\",\n \"Male\",\n \"COVID-19\",\n \"Viral Vaccines\",\n \"SARS-CoV-2\",\n \"Vaccination\",\n \"Hospitalization\"\n]"},"pmcid":{"kind":"string","value":"9645422"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Despite widespread COVID-19 vaccination, high community levels of SARS-CoV-2 circulating throughout the United States have led to many breakthrough SARS-CoV-2 infections [1-3]. Breakthrough infections, where fully vaccinated individuals who are exposed to SARS-CoV-2 get infected, are generally uncommon (0.02% of fully vaccinated individuals reported developing breakthrough infections in a Washington state cohort [4]) and are generally less severe than infections in unvaccinated individuals [5,6]. There now exists a large body of literature studying the risk factors for severe COVID-19 disease, much of which has involved studies in unvaccinated populations [7-10], prior to the large-scale availability of vaccines. Studies on how these risks vary after vaccination are fewer in comparison, mostly focused on vaccine effectiveness in preventing SARS-CoV-2 infections or the influence of specific variants on vaccine effectiveness [11].\nThe impact of underlying factors on breakthrough infections are quite challenging to understand outside of randomized, placebo-controlled, double-blind field trials due to variation in their severity, distribution in the population, and contribution to transmission [11,12]. Early studies have found that a third dose of vaccine reduces the viral load in breakthrough infections, even for newer variants such as delta and omicron [13,14]. To understand the comparative advantages of the various vaccines [15], it is important to know the rate of severe COVID-19 disease leading to hospitalization or death among individuals who are fully vaccinated [16], as this will help policy makers.\nWhile the risk of breakthrough SARS-CoV-2 infection has recently been reported by type of vaccine [17], little information exists regarding the risk of hospitalization or mortality by vaccine type for breakthrough infections [1]. In addition, while a prior SARS-CoV-2 infection is associated with a lower risk of breakthrough infection, it is unknown how large an effect a prior infection has on the severity of breakthrough COVID-19 infections, should one occur [18]. There has been a growing need for retrospective studies on severe breakthrough infections to address the misinformation and vaccine hesitancy in social and public spheres [19].\nIn this paper, we used de-identified US medical claims records from Change Healthcare to estimate the risk of hospitalization and death, by vaccine type, age, sex, comorbidity factors and previous SARS-CoV2 infection, among SARS-CoV-2 breakthrough infections that occurred between March 10, 2021, and October 14, 2021."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Ethics Approval and Consent to Participate [SUBSECTION] This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\nThis study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\n[SUBTITLE] Data Source [SUBSECTION] Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\nOur study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\n[SUBTITLE] Study Population [SUBSECTION] Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\nOur data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\n[SUBTITLE] Hospitalization and Mortality [SUBSECTION] We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\nWe explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\n[SUBTITLE] Study Period [SUBSECTION] COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\nCOVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\n[SUBTITLE] Comorbidities [SUBSECTION] Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\nPreexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\n[SUBTITLE] Previous COVID-19 Infection [SUBSECTION] Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\nSince some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\n[SUBTITLE] Statistical Methods [SUBSECTION] Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\nDate of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21]."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"Our study includes 19,815 fully vaccinated patients with breakthrough SARS-CoV-2 infections between March 10, 2021, and October 14, 2021. Of those patients, 11,339 (57.22%) received the Pfizer vaccine, 5480 (27.66%) received the Moderna vaccine, and 2996 (15.12%) received the Janssen vaccine. Breakthrough cases receiving Janssen were younger than those receiving Pfizer or Moderna and had a slightly greater proportion of male patients (Table 1). Breakthrough cases receiving Moderna had a greater proportion of patients with COVID-19 prior to vaccination.\nRisk of hospitalization and mortality among breakthrough cases increased with older age and was higher for male patients (Table 2). In multivariable analyses controlling for age, male sex, comorbidities, and prior SARS-CoV-2 infection, the risk of hospitalization was the lowest for breakthrough cases receiving the Moderna vaccine (adjusted hazard ratio [aHR]: 0.42, 95% CI 0.35-0.5; P<.001), comparably low for Pfizer vaccinated individuals (aHR: 0.45, 95% CI 0.39-0.53; P<.001), compared with that for the recipients of the Janssen vaccine. The comorbidities with statistically significant HRs for hospitalization or mortality from a breakthrough SARS-CoV-2 infection include severe liver disease, moderate and severe renal failures, alcohol abuse, chronic lung disease, coagulopathy, cancers, anemia, seizures, and arthritis (Table S1 in Multimedia Appendix 1).\nCharacteristics of SARS-CoV-2 breakthrough infections cohort tracked from March 10, 2021, to October 14, 2021. Prevalence of comorbidities by vaccine type is shown in Figure S5 in Multimedia Appendix 1.\nCorrelates of hospitalization and mortality after breakthrough SARS-CoV-2 infection, estimated from Cox proportional hazards models. We show the adjusted hazard ratio (aHR) and the 95% CI for the significant correlates (P values indicated via superscripts d, e, and f). An aHR of <1.0 indicates a lower risk of hospitalization or mortality as compared to the baseline population for that covariate (analogously, aHR>1.0 indicates a higher risk than the baseline). Hazard ratios (HRs) of comorbidities are shown in Table S1 in Multimedia Appendix 1.\naIncidence per 100 person years.\nbn=19,815; events=1140.\ncn=19,815; events=159.\ndP<.001.\neP<.05.\nfP<.01.\nWe see a similar trend with the risk of mortality for breakthrough cases, with the risk being the lowest for those receiving the Moderna vaccines (aHR: 0.38, 95% CI 0.23-0.62; P<.001) and comparably lower for Pfizer recipients (aHR: 0.43, 95% CI 0.28-0.65; P<.001) as compared to that for Janssen recipients. Finally, as expected, the protection offered by vaccines was enhanced for breakthrough cases who already had a previous SARS-CoV-2 infection. These individuals were 40% less likely to be hospitalized due to COVID-19 (aHR: 0.57, 95% CI 0.41-0.80; P=.001) and four times less likely to die of COVID-19 (aHR: 0.22, 95% CI 0.05-0.91; P=.04), when compared to those without a prior SARS-CoV-2 infection independent of age, sex, comorbidities, and vaccine type.\nWe repeat this analysis by excluding the population who had a prior SARS-CoV-2 infection for completeness and show the resulting HRs Table S3 in Multimedia Appendix 1."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Ethics Approval and Consent to Participate","Data Source","Study Population","Hospitalization and Mortality","Study Period","Comorbidities","Previous COVID-19 Infection","Statistical Methods","Principal Findings","Limitations","Conclusions"],"string":"[\n \"Ethics Approval and Consent to Participate\",\n \"Data Source\",\n \"Study Population\",\n \"Hospitalization and Mortality\",\n \"Study Period\",\n \"Comorbidities\",\n \"Previous COVID-19 Infection\",\n \"Statistical Methods\",\n \"Principal Findings\",\n \"Limitations\",\n \"Conclusions\"\n]"},"other_sections_texts":{"kind":"list like","value":["This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.","Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.","Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.","We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.","COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.","Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.","Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.","Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].","Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.","Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.","Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen."],"string":"[\n \"This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\",\n \"Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\",\n \"Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\",\n \"We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\",\n \"COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\",\n \"Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\",\n \"Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\",\n \"Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\",\n \"Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\",\n \"Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\",\n \"Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Ethics Approval and Consent to Participate","Data Source","Study Population","Hospitalization and Mortality","Study Period","Comorbidities","Previous COVID-19 Infection","Statistical Methods","Results","Discussion","Principal Findings","Limitations","Conclusions"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Ethics Approval and Consent to Participate\",\n \"Data Source\",\n \"Study Population\",\n \"Hospitalization and Mortality\",\n \"Study Period\",\n \"Comorbidities\",\n \"Previous COVID-19 Infection\",\n \"Statistical Methods\",\n \"Results\",\n \"Discussion\",\n \"Principal Findings\",\n \"Limitations\",\n \"Conclusions\"\n]"},"all_sections_texts":{"kind":"list like","value":["Despite widespread COVID-19 vaccination, high community levels of SARS-CoV-2 circulating throughout the United States have led to many breakthrough SARS-CoV-2 infections [1-3]. Breakthrough infections, where fully vaccinated individuals who are exposed to SARS-CoV-2 get infected, are generally uncommon (0.02% of fully vaccinated individuals reported developing breakthrough infections in a Washington state cohort [4]) and are generally less severe than infections in unvaccinated individuals [5,6]. There now exists a large body of literature studying the risk factors for severe COVID-19 disease, much of which has involved studies in unvaccinated populations [7-10], prior to the large-scale availability of vaccines. Studies on how these risks vary after vaccination are fewer in comparison, mostly focused on vaccine effectiveness in preventing SARS-CoV-2 infections or the influence of specific variants on vaccine effectiveness [11].\nThe impact of underlying factors on breakthrough infections are quite challenging to understand outside of randomized, placebo-controlled, double-blind field trials due to variation in their severity, distribution in the population, and contribution to transmission [11,12]. Early studies have found that a third dose of vaccine reduces the viral load in breakthrough infections, even for newer variants such as delta and omicron [13,14]. To understand the comparative advantages of the various vaccines [15], it is important to know the rate of severe COVID-19 disease leading to hospitalization or death among individuals who are fully vaccinated [16], as this will help policy makers.\nWhile the risk of breakthrough SARS-CoV-2 infection has recently been reported by type of vaccine [17], little information exists regarding the risk of hospitalization or mortality by vaccine type for breakthrough infections [1]. In addition, while a prior SARS-CoV-2 infection is associated with a lower risk of breakthrough infection, it is unknown how large an effect a prior infection has on the severity of breakthrough COVID-19 infections, should one occur [18]. There has been a growing need for retrospective studies on severe breakthrough infections to address the misinformation and vaccine hesitancy in social and public spheres [19].\nIn this paper, we used de-identified US medical claims records from Change Healthcare to estimate the risk of hospitalization and death, by vaccine type, age, sex, comorbidity factors and previous SARS-CoV2 infection, among SARS-CoV-2 breakthrough infections that occurred between March 10, 2021, and October 14, 2021.","[SUBTITLE] Ethics Approval and Consent to Participate [SUBSECTION] This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\nThis study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\n[SUBTITLE] Data Source [SUBSECTION] Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\nOur study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\n[SUBTITLE] Study Population [SUBSECTION] Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\nOur data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\n[SUBTITLE] Hospitalization and Mortality [SUBSECTION] We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\nWe explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\n[SUBTITLE] Study Period [SUBSECTION] COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\nCOVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\n[SUBTITLE] Comorbidities [SUBSECTION] Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\nPreexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\n[SUBTITLE] Previous COVID-19 Infection [SUBSECTION] Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\nSince some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\n[SUBTITLE] Statistical Methods [SUBSECTION] Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\nDate of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].","This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.","Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.","Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.","We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.","COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.","Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.","Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.","Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].","Our study includes 19,815 fully vaccinated patients with breakthrough SARS-CoV-2 infections between March 10, 2021, and October 14, 2021. Of those patients, 11,339 (57.22%) received the Pfizer vaccine, 5480 (27.66%) received the Moderna vaccine, and 2996 (15.12%) received the Janssen vaccine. Breakthrough cases receiving Janssen were younger than those receiving Pfizer or Moderna and had a slightly greater proportion of male patients (Table 1). Breakthrough cases receiving Moderna had a greater proportion of patients with COVID-19 prior to vaccination.\nRisk of hospitalization and mortality among breakthrough cases increased with older age and was higher for male patients (Table 2). In multivariable analyses controlling for age, male sex, comorbidities, and prior SARS-CoV-2 infection, the risk of hospitalization was the lowest for breakthrough cases receiving the Moderna vaccine (adjusted hazard ratio [aHR]: 0.42, 95% CI 0.35-0.5; P<.001), comparably low for Pfizer vaccinated individuals (aHR: 0.45, 95% CI 0.39-0.53; P<.001), compared with that for the recipients of the Janssen vaccine. The comorbidities with statistically significant HRs for hospitalization or mortality from a breakthrough SARS-CoV-2 infection include severe liver disease, moderate and severe renal failures, alcohol abuse, chronic lung disease, coagulopathy, cancers, anemia, seizures, and arthritis (Table S1 in Multimedia Appendix 1).\nCharacteristics of SARS-CoV-2 breakthrough infections cohort tracked from March 10, 2021, to October 14, 2021. Prevalence of comorbidities by vaccine type is shown in Figure S5 in Multimedia Appendix 1.\nCorrelates of hospitalization and mortality after breakthrough SARS-CoV-2 infection, estimated from Cox proportional hazards models. We show the adjusted hazard ratio (aHR) and the 95% CI for the significant correlates (P values indicated via superscripts d, e, and f). An aHR of <1.0 indicates a lower risk of hospitalization or mortality as compared to the baseline population for that covariate (analogously, aHR>1.0 indicates a higher risk than the baseline). Hazard ratios (HRs) of comorbidities are shown in Table S1 in Multimedia Appendix 1.\naIncidence per 100 person years.\nbn=19,815; events=1140.\ncn=19,815; events=159.\ndP<.001.\neP<.05.\nfP<.01.\nWe see a similar trend with the risk of mortality for breakthrough cases, with the risk being the lowest for those receiving the Moderna vaccines (aHR: 0.38, 95% CI 0.23-0.62; P<.001) and comparably lower for Pfizer recipients (aHR: 0.43, 95% CI 0.28-0.65; P<.001) as compared to that for Janssen recipients. Finally, as expected, the protection offered by vaccines was enhanced for breakthrough cases who already had a previous SARS-CoV-2 infection. These individuals were 40% less likely to be hospitalized due to COVID-19 (aHR: 0.57, 95% CI 0.41-0.80; P=.001) and four times less likely to die of COVID-19 (aHR: 0.22, 95% CI 0.05-0.91; P=.04), when compared to those without a prior SARS-CoV-2 infection independent of age, sex, comorbidities, and vaccine type.\nWe repeat this analysis by excluding the population who had a prior SARS-CoV-2 infection for completeness and show the resulting HRs Table S3 in Multimedia Appendix 1.","[SUBTITLE] Principal Findings [SUBSECTION] Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\nUsing medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\n[SUBTITLE] Limitations [SUBSECTION] Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\nLimitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\n[SUBTITLE] Conclusions [SUBSECTION] Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.\nOur findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.","Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.","Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.","Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen."],"string":"[\n \"Despite widespread COVID-19 vaccination, high community levels of SARS-CoV-2 circulating throughout the United States have led to many breakthrough SARS-CoV-2 infections [1-3]. Breakthrough infections, where fully vaccinated individuals who are exposed to SARS-CoV-2 get infected, are generally uncommon (0.02% of fully vaccinated individuals reported developing breakthrough infections in a Washington state cohort [4]) and are generally less severe than infections in unvaccinated individuals [5,6]. There now exists a large body of literature studying the risk factors for severe COVID-19 disease, much of which has involved studies in unvaccinated populations [7-10], prior to the large-scale availability of vaccines. Studies on how these risks vary after vaccination are fewer in comparison, mostly focused on vaccine effectiveness in preventing SARS-CoV-2 infections or the influence of specific variants on vaccine effectiveness [11].\\nThe impact of underlying factors on breakthrough infections are quite challenging to understand outside of randomized, placebo-controlled, double-blind field trials due to variation in their severity, distribution in the population, and contribution to transmission [11,12]. Early studies have found that a third dose of vaccine reduces the viral load in breakthrough infections, even for newer variants such as delta and omicron [13,14]. To understand the comparative advantages of the various vaccines [15], it is important to know the rate of severe COVID-19 disease leading to hospitalization or death among individuals who are fully vaccinated [16], as this will help policy makers.\\nWhile the risk of breakthrough SARS-CoV-2 infection has recently been reported by type of vaccine [17], little information exists regarding the risk of hospitalization or mortality by vaccine type for breakthrough infections [1]. In addition, while a prior SARS-CoV-2 infection is associated with a lower risk of breakthrough infection, it is unknown how large an effect a prior infection has on the severity of breakthrough COVID-19 infections, should one occur [18]. There has been a growing need for retrospective studies on severe breakthrough infections to address the misinformation and vaccine hesitancy in social and public spheres [19].\\nIn this paper, we used de-identified US medical claims records from Change Healthcare to estimate the risk of hospitalization and death, by vaccine type, age, sex, comorbidity factors and previous SARS-CoV2 infection, among SARS-CoV-2 breakthrough infections that occurred between March 10, 2021, and October 14, 2021.\",\n \"[SUBTITLE] Ethics Approval and Consent to Participate [SUBSECTION] This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\\nThis study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\\n[SUBTITLE] Data Source [SUBSECTION] Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\\nOur study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\\n[SUBTITLE] Study Population [SUBSECTION] Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\\nOur data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\\n[SUBTITLE] Hospitalization and Mortality [SUBSECTION] We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\\nWe explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\\n[SUBTITLE] Study Period [SUBSECTION] COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\\nCOVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\\n[SUBTITLE] Comorbidities [SUBSECTION] Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\\nPreexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\\n[SUBTITLE] Previous COVID-19 Infection [SUBSECTION] Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\\nSince some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\\n[SUBTITLE] Statistical Methods [SUBSECTION] Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\\nDate of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\",\n \"This study does not constitute as human subjects research due to the use and reporting of only deidentified observational data as determined by the human subjects committee of the University of Washington and thus does not require the review and approval by the institutional review board at the University of Washington.\",\n \"Our study uses de-identified US medical claims records from Change Healthcare collected over a period from March 1, 2020, to October 14, 2021, encompassing over 100 million records from over 8 million patients. Medical claims data contain details about a patient’s interaction with the medical system, which are needed for the accurate billing of the transactions. Each claims record contains patient demographic information, International Classification of Diseases, 10th Revision (ICD-10) codes indicating primary diagnosis and secondary diagnoses, place of diagnosis, ICD-10 codes of procedures performed, patient status at the end of the visit, dates pertaining to the event (where different “from” and “to” dates indicate longer visits whereas the same “from” and “to” dates are for outpatient visits).\\nOur claims data set includes primarily open claims and a subset of closed payer claims that are normalized for analytics purposes. The open claims are derived from broad-based health care sources and consist of all medical claims that Change Healthcare processes and for which it has the usage rights. The closed claims are derived directly from the payer (ie, health insurance provider) and capture nearly all events that occur during the patient’s enrollment period. Roughly 95% of the claims used for this study are commercial, and 5% are Medicare Advantage or other types of plans.\",\n \"Our data set of 8.18 million individuals contains only COVID-19 positive patients, defined as patients with at least one claims record with the ICD-10 diagnosis codes of “U07.1” or “U07.2” in any diagnosis field. We limited our analysis to individuals who had a primary diagnosis of “U07.1” (this is indicated by the principal diagnosis code, which encodes the primary diagnosis rendered by the medical facility or the primary cause of the visit). This ICD-10 diagnosis code indicates a COVID-19 diagnosis where the virus was identified in a lab-confirmed report. We exclude patients for whom the code of U07.1 appears in the “other diagnosis” fields, which contain the list of diagnoses made in addition to the primary cause of visit, which can be any other medical condition such as cancer. We also exclude patients with the code U07.2, which indicates a non-lab–confirmed COVID-19 diagnosis.\\nSubsequently, fully vaccinated individuals are identified by looking for procedure codes encoding the second doses of Pfizer (0002A) and Moderna (0012A) vaccines and the first dose of the Janssen (0031A) vaccine. We do not exclude patients with missing first dose claims records (~5% of the final study cohort), because patients who went to vaccination camps and were not required to provide insurance information would have missing first dose claims records. Some of these ~5% patients with missing first dose claims records may have had mixed vaccines (eg, Pfizer for the first dose and Moderna for the second dose). Since we did not believe that this will be a significant fraction of the vaccinated population, we do not exclude them. Breakthrough patients were defined as those who had a primary, lab-confirmed COVID-19 diagnosis at least 14 days after the date of vaccination. Please see Figure S1 in Multimedia Appendix 1 for a flow diagram showing the criteria used for cohort selection.\",\n \"We explicitly identify hospitalization by looking for claims where the claim type is “institutional” or “professional” and the bill type indicates an “inpatient” facility. We also look at the dates associated with the hospital stay and only consider patients whose admission duration was at least 2 days (derived from the “admission_from” and “admission_to” date fields). For mortality, we look at the patient status code and consider all codes indicating “expired.” As described already, we only consider cases where the primary diagnosis was COVID-19 for both hospitalization and expiration. The patient status code is available for all hospitalized patients but only for 42% of the outpatients (who went to clinics). Among patients who had the patient status code available, we found only 17 (0.22%) deaths out of a total of 7843 outpatients; therefore, we consider outpatients with missing patient status to be alive.\",\n \"COVID-19 vaccinations began in the United States in late December 2020. By late February 2021, the Pfizer-BioNTech (Pfizer), Moderna, and Johnson & Johnson (J&J/Janssen) vaccines were all approved for emergency use authorization. The Pfizer and Moderna vaccination drives started much earlier, in late December (Figures S2 and S3 in Multimedia Appendix 1), as compared to that of the Janssen vaccines, which also saw a stall in vaccine rates in mid-April (Figure S4 in Multimedia Appendix 1). To keep the COVID-19 exposure of the individuals taking any of the 3 vaccines consistent, we use the same study window, though we have data for Pfizer and Moderna from late December. We construct our cohort to consist of individuals who were fully vaccinated between March 10, 2021, and April 27, 2021, the period during which all 3 vaccines were being widely administered. Every individual in this cohort was followed from the date of vaccination of each individual up to the end of the study period, October 14, 2021. The study period over the entire cohort is thus March 10, 2021, to October 14, 2021. In Figure S6 in Multimedia Appendix 1, we show statistics showing the number of days of follow-up after full vaccination by vaccine type in our study cohort.\",\n \"Preexisting comorbidities were defined based on ICD-10 codes assigned to medical encounters, which contain pointers to previously diagnosed conditions, using claims records during the 6-month period from March 2020 to September 2020. This period does not overlap with the study period, so events during the study period will not also be counted as comorbidities. The Elixhauser comorbidity index [20] was used to define comorbid conditions. This index has a series of codes that define comorbidities with each code mapping to one or several ICD-10 diagnosis codes. For example, the Elixhauser code “BLDLOSS” (blood loss) includes the following four ICD-10 diagnosis codes: D50.0, O90.81, O99.02, and O99.03. We provide the index that we used and the corresponding ICD-10 codes in Multimedia Appendix 2. We also show the relative abundance of comorbidities in our cohort, by vaccine type, in Table S2 and Figure S5 in Multimedia Appendix 1.\",\n \"Since some of the individuals in our cohort may have had a COVID-19 infection during the year 2020, we introduce an additional feature to encode the effect of already being infected with COVID-19. This feature is “yes” if we see a claim involving a COVID-19 diagnosis in any diagnosis field, in the period from March 1, 2020, to the beginning of the study period, March 10, 2021.\",\n \"Date of full vaccination was defined as 14 days after (1) a single Janssen vaccine, (2) the second Moderna vaccine dose, or (3) the second Pfizer vaccine dose. Cox proportional hazards regression was used to estimate univariate hazard ratios (HRs) and multivariable HRs in a model including the following features: age (categorized), sex (male and female), vaccine type, Elixhauser comorbidities (encoded as independent binary variables), and SARS-CoV-2 infection prior to the first dose of vaccination (yes or no). We remove the negligible number of individuals with sex=unknown. We also model interactions between vaccine type and all other covariates as well as previous infection and all other covariates but find that none were statistically significant. Further, the interaction terms had a negligible impact on the hazard ratios of the other terms and were thus removed for greater clarity in the results. All analyses were performed using the “coxph” function from the R package “survival” (R Foundation for Statistical Computing) [21].\",\n \"Our study includes 19,815 fully vaccinated patients with breakthrough SARS-CoV-2 infections between March 10, 2021, and October 14, 2021. Of those patients, 11,339 (57.22%) received the Pfizer vaccine, 5480 (27.66%) received the Moderna vaccine, and 2996 (15.12%) received the Janssen vaccine. Breakthrough cases receiving Janssen were younger than those receiving Pfizer or Moderna and had a slightly greater proportion of male patients (Table 1). Breakthrough cases receiving Moderna had a greater proportion of patients with COVID-19 prior to vaccination.\\nRisk of hospitalization and mortality among breakthrough cases increased with older age and was higher for male patients (Table 2). In multivariable analyses controlling for age, male sex, comorbidities, and prior SARS-CoV-2 infection, the risk of hospitalization was the lowest for breakthrough cases receiving the Moderna vaccine (adjusted hazard ratio [aHR]: 0.42, 95% CI 0.35-0.5; P<.001), comparably low for Pfizer vaccinated individuals (aHR: 0.45, 95% CI 0.39-0.53; P<.001), compared with that for the recipients of the Janssen vaccine. The comorbidities with statistically significant HRs for hospitalization or mortality from a breakthrough SARS-CoV-2 infection include severe liver disease, moderate and severe renal failures, alcohol abuse, chronic lung disease, coagulopathy, cancers, anemia, seizures, and arthritis (Table S1 in Multimedia Appendix 1).\\nCharacteristics of SARS-CoV-2 breakthrough infections cohort tracked from March 10, 2021, to October 14, 2021. Prevalence of comorbidities by vaccine type is shown in Figure S5 in Multimedia Appendix 1.\\nCorrelates of hospitalization and mortality after breakthrough SARS-CoV-2 infection, estimated from Cox proportional hazards models. We show the adjusted hazard ratio (aHR) and the 95% CI for the significant correlates (P values indicated via superscripts d, e, and f). An aHR of <1.0 indicates a lower risk of hospitalization or mortality as compared to the baseline population for that covariate (analogously, aHR>1.0 indicates a higher risk than the baseline). Hazard ratios (HRs) of comorbidities are shown in Table S1 in Multimedia Appendix 1.\\naIncidence per 100 person years.\\nbn=19,815; events=1140.\\ncn=19,815; events=159.\\ndP<.001.\\neP<.05.\\nfP<.01.\\nWe see a similar trend with the risk of mortality for breakthrough cases, with the risk being the lowest for those receiving the Moderna vaccines (aHR: 0.38, 95% CI 0.23-0.62; P<.001) and comparably lower for Pfizer recipients (aHR: 0.43, 95% CI 0.28-0.65; P<.001) as compared to that for Janssen recipients. Finally, as expected, the protection offered by vaccines was enhanced for breakthrough cases who already had a previous SARS-CoV-2 infection. These individuals were 40% less likely to be hospitalized due to COVID-19 (aHR: 0.57, 95% CI 0.41-0.80; P=.001) and four times less likely to die of COVID-19 (aHR: 0.22, 95% CI 0.05-0.91; P=.04), when compared to those without a prior SARS-CoV-2 infection independent of age, sex, comorbidities, and vaccine type.\\nWe repeat this analysis by excluding the population who had a prior SARS-CoV-2 infection for completeness and show the resulting HRs Table S3 in Multimedia Appendix 1.\",\n \"[SUBTITLE] Principal Findings [SUBSECTION] Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\\nUsing medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\\n[SUBTITLE] Limitations [SUBSECTION] Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\\nLimitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\\n[SUBTITLE] Conclusions [SUBSECTION] Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.\\nOur findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.\",\n \"Using medical claims data, we found that the risk of hospitalization in SARS-CoV-2 breakthrough infections was lower for those receiving the Moderna and Pfizer vaccines compared to those receiving the Janssen vaccine. The risk of mortality was similarly low in breakthrough infections who received Pfizer and Moderna vaccines compared to those receiving the Janssen vaccine. There was no statistically significant difference between the HRs of Pfizer and Moderna for both risks. We also found older age, male sex, and certain comorbidities to be risk factors for hospitalization and mortality in breakthrough infections. Further, we found that risk of hospitalization was 40% less and risk of death was 75% less in SARS-CoV-2 breakthrough infections among individuals who already had a SARS-CoV-2 infection prior to their vaccination compared with fully vaccinated individuals without a previous SARS-CoV-2 infection. While other studies have reported lower risk of breakthrough infection with previous SARS-CoV-2 infection [18], our study analyzes both hospitalization and mortality and shows that the immunity provided by previous infection seems to increase the protection provided by vaccines, against severe COVID-19, independent of vaccine type, age, comorbidities, and sex. Since our cohort only consists of individuals who were all fully vaccinated, this is by no means a comparison of vaccine-induced immunity against acquired immunity from previous infections.\\nExcluding patients who had COVID-19 infection prior to vaccination from our Cox regression analysis results in a similar HR for hospitalization risk in patients who received the Pfizer (aHR=0.42) and Moderna (aHR=0.41) vaccines (Table S3 in Multimedia Appendix 1). This might be explained by the fact that 20.7% of patients who received Moderna had a prior COVID-19 infection as compared to ~13% of patients who received Pfizer. Hence, removing all patients with prior COVID-19 infection reduced the influence of the additional immunity that some of Moderna-vaccinated individuals had.\\nA number of studies have found that age has a direct effect on the risk of severe COVID-19 disease [22,23]. We find that the proportion of the elderly cohort in our data set who were hospitalized is much higher than the proportion of the younger cohort (Figure S7 in multimedia Appendix 1). In addition, we find a higher HR for the elderly subset of our study cohort (aHR=2.1 for age>50, aHR=3.3 for age>65, and aHR=5.0 for age>80, as compared to the baseline age group of 35-50 years).\\nOur findings comparing vaccine types are similar to those reported by the Centers for Disease Control and Prevention for mortality but provide additional information by vaccine type [1,16]. There have been several studies on individual risk factors such as age [22], specific comorbidities [7], focused populations such as Veterans [24], or large-scale projects such as OpenSAFELY, which involved 17 million unvaccinated patients [10]. Our work advances this body of literature by analyzing vaccine type, age, sex, and 39 different comorbidities in a large cohort of breakthrough patients from the general US population. Some of the risk factors that we find for severe breakthrough SARS-CoV-2 infections, such as age, male gender, and certain comorbidities (eg, chronic lung infection, kidney disease, and cancers) are similar to what has been reported in prior studies of SARS-CoV-2 infections among unvaccinated individuals [8,23,25]. However, we find that some risk factors found by initial studies such as hypertension are not a risk factor for breakthrough COVID-19 hospitalization or death (aHRs of 0.75 and 0.59, respectively), neither are diabetes or congestive heart failure. We instead find that both moderate and severe renal failure are significant risk factors, independent of age or other factors, which agrees with other large-scale studies such as OpenSAFELY [10] and the Global Burden of Disease collaboration [26] which identified that worldwide chronic kidney disease is the most prevalent risk factor for severe COVID-19. Even mild impairment of renal function has been found to be an independent risk factor for COVID-19 infection, hospitalization, and mortality [27].\\nLastly, to understand the association between outcomes and the time of vaccination, we incorporate a variable indicating the number of days between full vaccination and the onset of the surge in infections caused by the delta variant. However, our population-based data set is inadequate to derive any significant conclusions vis-à-vis the best time for vaccination in anticipation of a surge.\",\n \"Limitations of our study include, first, a lack of access to data on unvaccinated individuals or those who had a negative SARS-CoV-2 test result. The former is due to the lack of a medical claims record for vaccinations that were done in vaccination drives and camps; the absence of a vaccination-related claim in our data set therefore does not imply an unvaccinated individual. Second, our medical claims source consists of mostly privately insured individuals and can thus miss people who may be susceptible to the most adverse outcomes. Another caveat of our data set is that most of the claims are open claims, which have the benefit of capturing a patient’s activities over a longer time frame regardless of their insurance provider, but do not necessarily track all medical encounters of patients.\",\n \"Our findings add to the growing literature regarding the risk factors for severe breakthrough SARS-CoV-2 infections in fully vaccinated individuals, where we identify the influence of age, sex, and comorbidities that are risk factors; importantly, we found that previous SARS-CoV-2 infections can provide additional protection over that offered by vaccines against severe disease. Our results also necessitate further studies on the optimal number of vaccine doses to protect from the most severe breakthrough SARS-CoV-2 infections. An important strength of our study is that we consider US-wide breakthrough hospitalizations covering a broad demographic and compare all 3 vaccines, whereas most previous studies lack specific data on Janssen.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction","methods",null,null,null,null,null,null,null,null,"results","discussion",null,null,null],"string":"[\n \"introduction\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n \"discussion\",\n null,\n null,\n null\n]"},"keywords":{"kind":"list like","value":["breakthroughs","vaccines","Pfizer","Moderna","Janssen","SARS-CoV-2","COVID-19","coronavirus","infectious disease","viral infection","vaccination","breakthrough infection","public health","health policy","decision making","booster vaccine","mortality","hospitalization","healthcare system"],"string":"[\n \"breakthroughs\",\n \"vaccines\",\n \"Pfizer\",\n \"Moderna\",\n \"Janssen\",\n \"SARS-CoV-2\",\n \"COVID-19\",\n \"coronavirus\",\n \"infectious disease\",\n \"viral infection\",\n \"vaccination\",\n \"breakthrough infection\",\n \"public health\",\n \"health policy\",\n \"decision making\",\n \"booster vaccine\",\n \"mortality\",\n \"hospitalization\",\n \"healthcare system\"\n]"}}},{"rowIdx":368,"cells":{"title":{"kind":"string","value":"Outcomes of COVID-19 Infection in People Previously Vaccinated Against Influenza: Population-Based Cohort Study Using Primary Health Care Electronic Records."},"pmid":{"kind":"string","value":"36265160"},"background_abstract":{"kind":"string","value":"A possible link between influenza immunization and susceptibility to the complications of COVID-19 infection has been previously suggested owing to a boost in the immunity against SARS-CoV-2."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We performed a population-based cohort study including all patients with COVID-19 with registered entries in the primary health care (PHC) electronic records during the first wave of the COVID-19 pandemic (March 1 to June 30, 2020) in Catalonia, Spain. We compared individuals who took an influenza vaccine before being infected with COVID-19, with those who had not taken one. Data were obtained from Information System for Research in Primary Care, capturing PHC information of 5.8 million people from Catalonia. The main outcomes assessed during follow-up were a diagnosis of pneumonia, hospital admission, and mortality."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"We included 309,039 individuals with COVID-19 and compared them on the basis of their influenza immunization status, with 114,181 (36.9%) having been vaccinated at least once and 194,858 (63.1%) having never been vaccinated. In total, 21,721 (19%) vaccinated individuals and 11,000 (5.7%) unvaccinated individuals had at least one of their outcomes assessed. Those vaccinated against influenza at any time (odds ratio [OR] 1.14, 95% CI 1.10-1.19), recently (OR 1.13, 95% CI 1.10-1.18), or recurrently (OR 1.10, 95% CI 1.05-1.15) before being infected with COVID-19 had a higher risk of presenting at least one of the outcomes than did unvaccinated individuals. When we excluded people living in long-term care facilities, the results were similar."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"We could not establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection, as the risk of COVID-19 complications was higher in vaccinated than in unvaccinated individuals. Our results correspond to the first wave of the COVID-19 pandemic, where more complications and mortalities due to COVID-19 had occurred. Despite that, our study adds more evidence for the analysis of a possible link between the quality of immunity and COVID-19 outcomes, particularly in the PHC setting."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Influenza Vaccines","Influenza, Human","COVID-19","Pandemics","SARS-CoV-2","Cohort Studies","Primary Health Care","Electronics"],"string":"[\n \"Humans\",\n \"Influenza Vaccines\",\n \"Influenza, Human\",\n \"COVID-19\",\n \"Pandemics\",\n \"SARS-CoV-2\",\n \"Cohort Studies\",\n \"Primary Health Care\",\n \"Electronics\"\n]"},"pmcid":{"kind":"string","value":"9662290"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"COVID-19 is caused by SARS-CoV-2, a novel coronavirus that emerged in China in 2019, which became the primary agent of a new pandemic that rapidly spread worldwide [1], with an average global infection fatality rate of approximately 0.15%, depending on the data analyzed [2]. SARS-CoV-2 mainly affects the respiratory tract and uses surface proteins in order to infect the host [3].\nAlthough new variants of SARS-CoV-2 have emerged since December 2020, the coronavirus’ genome is composed of RNA and depends on the RNA polymerase to generate its proteins, with a mechanism of error correction that results in a lower mutation rate than the influenza virus [4]. This low mutation rate may suggest that the vaccines developed against SARS-CoV-2, as well as the immunity generated in those patients who were infected, could represent a long-lasting immunity [5,6].\nCOVID-19, similar to influenza A and B, is caused by RNA virus and produce similar symptoms. The influenza virus needs the hemagglutinin and neuraminidase surface proteins to infect cells, whereas SARS-CoV-2 needs the S protein [5]. Previous in vitro and animal studies suggest an induction pathway of indirect etiological immunity between the influenza vaccine and SARS-CoV-2. Animal models suggest that some influenza subtypes might lead to regulation of the angiotensin-converting enzyme-2, with protective properties against SARS [7]. An unspecific effect of infection and vaccination on the immune system and susceptibility to other infections has also been reported, albeit with discordant data [8-10]. Some modeling studies have suggested a possible association between influenza immunization and COVID-19 [11-14].\nA study conducted in Australia assessed the cellular and humoral immune responses during and after disease occurrence in a patient with a mild COVID-19 infection. They found that the immune response in different cell types is associated with clinical recovery. These results are coincident with similar findings among patients with influenza reported by the same authors [15,16].\nOther studies observed differences in the susceptibility to COVID-19 in children of different ages with a lower infection rate than that in adults and older individuals [17]. Although the mechanism underlying these differences in severity and susceptibility is unclear, a possible explanation might be the difference in the quantity and quality of the immune function determined by the history of infections and the recent vaccines administered [18].\nConsequently, a link between the quality of the immunity and recovery from COVID-19 may exist. Thus, we hypothesized that the immunity resulting from the previous influenza vaccination would boost part of the immunity against SARS-CoV-2, and we aimed to investigate whether individuals with COVID-19 could have benefited from vaccination against influenza."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Study Design [SUBSECTION] We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\nWe performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\n[SUBTITLE] Data Source [SUBSECTION] Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\nOur data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\n[SUBTITLE] COVID-19 Classification [SUBSECTION] Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\nParticipants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\n[SUBTITLE] Influenza Immunization [SUBSECTION] Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\nPatients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\n[SUBTITLE] Variables [SUBSECTION] At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\nAt baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\n[SUBTITLE] Statistical Analysis [SUBSECTION] Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\nQuantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\n[SUBTITLE] Ethical Considerations [SUBSECTION] The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\nThe study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679)."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"We included 309,039 individuals with COVID-19 during the first wave of the pandemic in accordance with their influenza immunization status (Table 1, Multimedia Appendix 1); 114,181 (36.9%) participants had received the influenza vaccine at least once before having COVID-19 and 194,858 (63.1%) had not been vaccinated, with more women in both groups, especially in the vaccinated cohort (61.0% women vs 39.0% men). The mean age was higher for vaccinated individuals (64.3 years, with 52.3% of them being older than 65 years). Vaccinated individuals had more comorbidities than unvaccinated individuals.\nOf those who received the influenza vaccine, 66,611 (58.3%) had been recently vaccinated (2019-2020) and 75,311 (66%) had been systematically vaccinated against influenza at least during 3 different years (Table 2).\nOf the participants with COVID-19, 11,000 (5.7%) unvaccinated and 21,721 (19%) vaccinated participants presented at least one of the following events: hospital admission, pneumonia, or death. For those who received the influenza vaccine at any time before having COVID-19, the risks of hospitalization (adjusted OR 1.14, 95% CI 1.10-1.19) and death (OR 1.32, 1.23-1.42) were higher than those among unvaccinated participants. For the recently vaccinated participants, the risk was higher for hospitalization (OR 1.16, 95% CI 1.1-1.23), pneumonia (OR 1.12, 95% CI 1.02-1.23), and death (OR 1.14, 95% CI 1.04-1.24). For people with recurrent vaccination, the risk of the 3 outcomes was also higher that among unvaccinated participants (OR 1.07, 1.16, and 1.24, respectively; Table 3). We have also analyzed the results in a matched population of vaccinated versus unvaccinated participants, revealing a higher risk of pneumonia and mortality, with an adjusted OR of 1.11 (95% CI 1.01-1.23) and 1.28 (95% CI 1.07-1.53), respectively (Multimedia Appendix 2).\nThe risks of the outcomes based on influenza vaccination status and excluding those patients living in LTCFs are shown in Figure 1. For non-LTCF residents, the results are similar to those for the whole population, except that there was no significant increase in mortality (OR 0.93, 95% CI 0.85-1.03).\nSociodemographic and clinical characteristics of the study population (N=309,039).\nTaking influenza vaccines before having COVID-19.\nLogistic regression model of COVID-19 outcomes based on influenza immunization status.\naA logistic regression model adjusted with the following relevant covariables was fitted: smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), co-medication, and previous vaccines (pneumococcal and tuberculosis).\nRisk of death and of combined COVID-19 complications in all the vaccinated population and excluding people living in long-term care facilities (LTCF)."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Study Design","Data Source","COVID-19 Classification","Influenza Immunization","Variables","Statistical Analysis","Ethical Considerations","Principal Findings","Comparison With Prior Work","Limitations","Conclusions"],"string":"[\n \"Study Design\",\n \"Data Source\",\n \"COVID-19 Classification\",\n \"Influenza Immunization\",\n \"Variables\",\n \"Statistical Analysis\",\n \"Ethical Considerations\",\n \"Principal Findings\",\n \"Comparison With Prior Work\",\n \"Limitations\",\n \"Conclusions\"\n]"},"other_sections_texts":{"kind":"list like","value":["We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.","Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].","Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].","Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].","At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].","Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).","The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).","We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.","Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.","We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.","In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies."],"string":"[\n \"We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\",\n \"Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\",\n \"Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\",\n \"Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\",\n \"At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\",\n \"Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\",\n \"The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\",\n \"We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\",\n \"Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\",\n \"We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\",\n \"In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Study Design","Data Source","COVID-19 Classification","Influenza Immunization","Variables","Statistical Analysis","Ethical Considerations","Results","Discussion","Principal Findings","Comparison With Prior Work","Limitations","Conclusions"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Study Design\",\n \"Data Source\",\n \"COVID-19 Classification\",\n \"Influenza Immunization\",\n \"Variables\",\n \"Statistical Analysis\",\n \"Ethical Considerations\",\n \"Results\",\n \"Discussion\",\n \"Principal Findings\",\n \"Comparison With Prior Work\",\n \"Limitations\",\n \"Conclusions\"\n]"},"all_sections_texts":{"kind":"list like","value":["COVID-19 is caused by SARS-CoV-2, a novel coronavirus that emerged in China in 2019, which became the primary agent of a new pandemic that rapidly spread worldwide [1], with an average global infection fatality rate of approximately 0.15%, depending on the data analyzed [2]. SARS-CoV-2 mainly affects the respiratory tract and uses surface proteins in order to infect the host [3].\nAlthough new variants of SARS-CoV-2 have emerged since December 2020, the coronavirus’ genome is composed of RNA and depends on the RNA polymerase to generate its proteins, with a mechanism of error correction that results in a lower mutation rate than the influenza virus [4]. This low mutation rate may suggest that the vaccines developed against SARS-CoV-2, as well as the immunity generated in those patients who were infected, could represent a long-lasting immunity [5,6].\nCOVID-19, similar to influenza A and B, is caused by RNA virus and produce similar symptoms. The influenza virus needs the hemagglutinin and neuraminidase surface proteins to infect cells, whereas SARS-CoV-2 needs the S protein [5]. Previous in vitro and animal studies suggest an induction pathway of indirect etiological immunity between the influenza vaccine and SARS-CoV-2. Animal models suggest that some influenza subtypes might lead to regulation of the angiotensin-converting enzyme-2, with protective properties against SARS [7]. An unspecific effect of infection and vaccination on the immune system and susceptibility to other infections has also been reported, albeit with discordant data [8-10]. Some modeling studies have suggested a possible association between influenza immunization and COVID-19 [11-14].\nA study conducted in Australia assessed the cellular and humoral immune responses during and after disease occurrence in a patient with a mild COVID-19 infection. They found that the immune response in different cell types is associated with clinical recovery. These results are coincident with similar findings among patients with influenza reported by the same authors [15,16].\nOther studies observed differences in the susceptibility to COVID-19 in children of different ages with a lower infection rate than that in adults and older individuals [17]. Although the mechanism underlying these differences in severity and susceptibility is unclear, a possible explanation might be the difference in the quantity and quality of the immune function determined by the history of infections and the recent vaccines administered [18].\nConsequently, a link between the quality of the immunity and recovery from COVID-19 may exist. Thus, we hypothesized that the immunity resulting from the previous influenza vaccination would boost part of the immunity against SARS-CoV-2, and we aimed to investigate whether individuals with COVID-19 could have benefited from vaccination against influenza.","[SUBTITLE] Study Design [SUBSECTION] We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\nWe performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\n[SUBTITLE] Data Source [SUBSECTION] Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\nOur data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\n[SUBTITLE] COVID-19 Classification [SUBSECTION] Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\nParticipants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\n[SUBTITLE] Influenza Immunization [SUBSECTION] Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\nPatients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\n[SUBTITLE] Variables [SUBSECTION] At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\nAt baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\n[SUBTITLE] Statistical Analysis [SUBSECTION] Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\nQuantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\n[SUBTITLE] Ethical Considerations [SUBSECTION] The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\nThe study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).","We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.","Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].","Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].","Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].","At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].","Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).","The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).","We included 309,039 individuals with COVID-19 during the first wave of the pandemic in accordance with their influenza immunization status (Table 1, Multimedia Appendix 1); 114,181 (36.9%) participants had received the influenza vaccine at least once before having COVID-19 and 194,858 (63.1%) had not been vaccinated, with more women in both groups, especially in the vaccinated cohort (61.0% women vs 39.0% men). The mean age was higher for vaccinated individuals (64.3 years, with 52.3% of them being older than 65 years). Vaccinated individuals had more comorbidities than unvaccinated individuals.\nOf those who received the influenza vaccine, 66,611 (58.3%) had been recently vaccinated (2019-2020) and 75,311 (66%) had been systematically vaccinated against influenza at least during 3 different years (Table 2).\nOf the participants with COVID-19, 11,000 (5.7%) unvaccinated and 21,721 (19%) vaccinated participants presented at least one of the following events: hospital admission, pneumonia, or death. For those who received the influenza vaccine at any time before having COVID-19, the risks of hospitalization (adjusted OR 1.14, 95% CI 1.10-1.19) and death (OR 1.32, 1.23-1.42) were higher than those among unvaccinated participants. For the recently vaccinated participants, the risk was higher for hospitalization (OR 1.16, 95% CI 1.1-1.23), pneumonia (OR 1.12, 95% CI 1.02-1.23), and death (OR 1.14, 95% CI 1.04-1.24). For people with recurrent vaccination, the risk of the 3 outcomes was also higher that among unvaccinated participants (OR 1.07, 1.16, and 1.24, respectively; Table 3). We have also analyzed the results in a matched population of vaccinated versus unvaccinated participants, revealing a higher risk of pneumonia and mortality, with an adjusted OR of 1.11 (95% CI 1.01-1.23) and 1.28 (95% CI 1.07-1.53), respectively (Multimedia Appendix 2).\nThe risks of the outcomes based on influenza vaccination status and excluding those patients living in LTCFs are shown in Figure 1. For non-LTCF residents, the results are similar to those for the whole population, except that there was no significant increase in mortality (OR 0.93, 95% CI 0.85-1.03).\nSociodemographic and clinical characteristics of the study population (N=309,039).\nTaking influenza vaccines before having COVID-19.\nLogistic regression model of COVID-19 outcomes based on influenza immunization status.\naA logistic regression model adjusted with the following relevant covariables was fitted: smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), co-medication, and previous vaccines (pneumococcal and tuberculosis).\nRisk of death and of combined COVID-19 complications in all the vaccinated population and excluding people living in long-term care facilities (LTCF).","[SUBTITLE] Principal Findings [SUBSECTION] We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\nWe analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\n[SUBTITLE] Comparison With Prior Work [SUBSECTION] Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\nSome researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\n[SUBTITLE] Limitations [SUBSECTION] We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\nWe need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\n[SUBTITLE] Conclusions [SUBSECTION] In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.\nIn conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.","We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.","Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.","We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.","In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies."],"string":"[\n \"COVID-19 is caused by SARS-CoV-2, a novel coronavirus that emerged in China in 2019, which became the primary agent of a new pandemic that rapidly spread worldwide [1], with an average global infection fatality rate of approximately 0.15%, depending on the data analyzed [2]. SARS-CoV-2 mainly affects the respiratory tract and uses surface proteins in order to infect the host [3].\\nAlthough new variants of SARS-CoV-2 have emerged since December 2020, the coronavirus’ genome is composed of RNA and depends on the RNA polymerase to generate its proteins, with a mechanism of error correction that results in a lower mutation rate than the influenza virus [4]. This low mutation rate may suggest that the vaccines developed against SARS-CoV-2, as well as the immunity generated in those patients who were infected, could represent a long-lasting immunity [5,6].\\nCOVID-19, similar to influenza A and B, is caused by RNA virus and produce similar symptoms. The influenza virus needs the hemagglutinin and neuraminidase surface proteins to infect cells, whereas SARS-CoV-2 needs the S protein [5]. Previous in vitro and animal studies suggest an induction pathway of indirect etiological immunity between the influenza vaccine and SARS-CoV-2. Animal models suggest that some influenza subtypes might lead to regulation of the angiotensin-converting enzyme-2, with protective properties against SARS [7]. An unspecific effect of infection and vaccination on the immune system and susceptibility to other infections has also been reported, albeit with discordant data [8-10]. Some modeling studies have suggested a possible association between influenza immunization and COVID-19 [11-14].\\nA study conducted in Australia assessed the cellular and humoral immune responses during and after disease occurrence in a patient with a mild COVID-19 infection. They found that the immune response in different cell types is associated with clinical recovery. These results are coincident with similar findings among patients with influenza reported by the same authors [15,16].\\nOther studies observed differences in the susceptibility to COVID-19 in children of different ages with a lower infection rate than that in adults and older individuals [17]. Although the mechanism underlying these differences in severity and susceptibility is unclear, a possible explanation might be the difference in the quantity and quality of the immune function determined by the history of infections and the recent vaccines administered [18].\\nConsequently, a link between the quality of the immunity and recovery from COVID-19 may exist. Thus, we hypothesized that the immunity resulting from the previous influenza vaccination would boost part of the immunity against SARS-CoV-2, and we aimed to investigate whether individuals with COVID-19 could have benefited from vaccination against influenza.\",\n \"[SUBTITLE] Study Design [SUBSECTION] We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\\nWe performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\\n[SUBTITLE] Data Source [SUBSECTION] Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\\nOur data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\\n[SUBTITLE] COVID-19 Classification [SUBSECTION] Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\\nParticipants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\\n[SUBTITLE] Influenza Immunization [SUBSECTION] Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\\nPatients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\\n[SUBTITLE] Variables [SUBSECTION] At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\\nAt baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\\n[SUBTITLE] Statistical Analysis [SUBSECTION] Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\\nQuantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\\n[SUBTITLE] Ethical Considerations [SUBSECTION] The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\\nThe study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\",\n \"We performed a population-based cohort study including all adults with COVID-19 in Catalonia, Spain, who were registered as confirmed cases (through the polymerase chain reaction [PCR]) or as probable cases (not confirmed through PCR but with International Classification of Diseases (ICD)-10 codes registered that are compatible with COVID-19) in the primary health care (PHC) system. All individuals with COVID-19 were diagnosed from the pandemic’s onset (March 2020) to June 30, 2020. Participants were compared on the basis of their influenza vaccination status between those having received the influenza vaccine before having COVID-19 (vaccinated in the previous influenza seasonal campaign in 2019-2020 or before) [19] with those who were not vaccinated.\",\n \"Our data source is the Information System for Research in Primary Care [20], which captures clinical information of approximately 5.8 million people from Catalonia, Spain (approximately 80% of the Catalan population). This information is pseudonymized, having originated from different data sources: (1) electronic health records in PHC system of the Catalan Health Institute, including sociodemographic characteristics, residents in nursing homes or long-term care facilities (LTCFs), comorbidities registered as ICD-10 codes [21], specialist referrals, clinical parameters, toxic habits, sickness leave, date of death, laboratory test data, and drug prescriptions issued in the PHC system, registered in accordance with the anatomical therapeutic chemical classification system [22]; (2) pharmacy invoice data corresponding to the PHC drug prescriptions; (3) database of diagnoses upon hospital discharge [23]; and (4) COVID-19 data from the Catalan Agency of Health Quality and Evaluation (AQuAS) [24].\",\n \"Participants were classified in accordance with the following criteria: confirmed cases are those with a confirmed COVID-19 diagnosis record, positive PCR outcome, or a positive serology test result. Those with an unconfirmed diagnosis or test (possible or unclear) along with any individual with a record of hospitalization, pneumonia, or death related to COVID-19 were considered probable cases. During the first wave of the COVID-19 pandemic in Catalonia, PCR tests were not routinely conducted for all individuals with compatible symptoms owing to the unavailability of laboratory kits to carry out the tests. Thus, we needed to capture those patients with a possible diagnosis of COVID-19, such as those admitted to hospital with pneumonia or other COVID-19 symptoms, who were not tested. We designed an algorithm to classify patients as “COVID possible” when a test result was unavailable along with registered entries from different databases: PCR tests or serology tests conducted in different settings, discharge diagnoses of pneumonia from Catalan hospitals or from emergency departments, and ICD-10 diagnoses related to COVID-19 coded in PHC. The date of COVID-19 diagnosis was set to be the first of all records used per patient. To guarantee that our algorithm is not far from the Catalan population, the resulting cohort was compared to the official COVID-19 cases reported by the AQuAS during the pandemic [24].\",\n \"Patients were classified as having taken the influenza vaccine if they had been vaccinated at any time before having COVID-19, and grouped in accordance with the seasonal vaccination campaign: the immediate previous campaign (2019-2020) or other vaccination campaigns (2018-2019 and before) [19,25].\",\n \"At baseline, the following variables were captured: sex, age, geographical area, MEDEA (Mortalidad en áreas pequeñas Españolas y Desigualdades Económicas y Ambientales [Mortality in small Spanish areas and economic and environmental inequalities]) socioeconomic index (deprivation index based on 5 indicators of socioeconomic position; it helps analyze health inequity, and higher the MEDEA socioeconomic index, worse the deprivation) [26], BMI, residence in nursing homes, smoking habits, comorbidities, and taking influenza vaccines and pneumococcal and tuberculosis vaccines.\\nThe main outcomes assessed during follow-up (up to June 2020) were at least one of the following variables: diagnosis of pneumonia, hospital admission, and mortality. The risk of these events was analyzed in those people who had been vaccinated against influenza at any time before having COVID-19, in those who were recently vaccinated (campaign of 2019-2020), and in those systematically vaccinated (who had been vaccinated at least during 3 different campaigns). We analyzed the same outcomes excluding those of people living in LTCFs, where vaccination is nearly universal in our country [27].\",\n \"Quantitative variables were described as mean (SD) values, whereas categorical variables were described as the proportion of vaccinated and unvaccinated individuals. Univariate analyses were based on the Student t test or chi-square test depending on the variable.\\nFor each outcome, we fitted a logistic regression model to estimate an odds ratio (OR) comparing the prevalence of each outcome among individuals given the influenza vaccine to that of unvaccinated individuals. The logistic model was fitted along with other covariables such as smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), concomitant drugs, and previous vaccines (pneumococcal and tuberculosis). As a sensitivity analysis, we conducted the same analysis on a matched population. Individuals vaccinated against influenza and unvaccinated controls were matched 1:2 in accordance with their age and gender at the time of infection or on an index date, and the reported ORs were obtained by fitting a conditional logistic regression model (clogit) accounting for matched pairs and adjusted using the same covariables as in the logistic model. We used the Wald test on the fitted coefficient to determine whether the log-odds were significantly different from 0 at a threshold of .05. All analyses were performed in R (version 4.1.0 or above; The R Foundation).\",\n \"The study protocol was approved by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària (June 3, 2020). This is a database research study that has been conducted in accordance with the tenets of the Declaration of Helsinki (Fortaleza, Brazil 2013) and does not require consent from the study participants for the purpose of publication. The need for consent was waived by the Research Ethics Committee of Institut Universitari d’Investigació en Atenció Primària as it is deemed unnecessary according to the European legislation (Regulation [EU] 2016/679).\",\n \"We included 309,039 individuals with COVID-19 during the first wave of the pandemic in accordance with their influenza immunization status (Table 1, Multimedia Appendix 1); 114,181 (36.9%) participants had received the influenza vaccine at least once before having COVID-19 and 194,858 (63.1%) had not been vaccinated, with more women in both groups, especially in the vaccinated cohort (61.0% women vs 39.0% men). The mean age was higher for vaccinated individuals (64.3 years, with 52.3% of them being older than 65 years). Vaccinated individuals had more comorbidities than unvaccinated individuals.\\nOf those who received the influenza vaccine, 66,611 (58.3%) had been recently vaccinated (2019-2020) and 75,311 (66%) had been systematically vaccinated against influenza at least during 3 different years (Table 2).\\nOf the participants with COVID-19, 11,000 (5.7%) unvaccinated and 21,721 (19%) vaccinated participants presented at least one of the following events: hospital admission, pneumonia, or death. For those who received the influenza vaccine at any time before having COVID-19, the risks of hospitalization (adjusted OR 1.14, 95% CI 1.10-1.19) and death (OR 1.32, 1.23-1.42) were higher than those among unvaccinated participants. For the recently vaccinated participants, the risk was higher for hospitalization (OR 1.16, 95% CI 1.1-1.23), pneumonia (OR 1.12, 95% CI 1.02-1.23), and death (OR 1.14, 95% CI 1.04-1.24). For people with recurrent vaccination, the risk of the 3 outcomes was also higher that among unvaccinated participants (OR 1.07, 1.16, and 1.24, respectively; Table 3). We have also analyzed the results in a matched population of vaccinated versus unvaccinated participants, revealing a higher risk of pneumonia and mortality, with an adjusted OR of 1.11 (95% CI 1.01-1.23) and 1.28 (95% CI 1.07-1.53), respectively (Multimedia Appendix 2).\\nThe risks of the outcomes based on influenza vaccination status and excluding those patients living in LTCFs are shown in Figure 1. For non-LTCF residents, the results are similar to those for the whole population, except that there was no significant increase in mortality (OR 0.93, 95% CI 0.85-1.03).\\nSociodemographic and clinical characteristics of the study population (N=309,039).\\nTaking influenza vaccines before having COVID-19.\\nLogistic regression model of COVID-19 outcomes based on influenza immunization status.\\naA logistic regression model adjusted with the following relevant covariables was fitted: smoking habits, age, comorbidities (asthma, autoimmune disorders, prior cerebrovascular disease, chronic kidney disease, chronic pulmonary obstructive disease, diabetes, heart failure, hypertension, ischemic heart disease, mental-behavioral disorders, obesity, organ transplant, and other respiratory diseases), co-medication, and previous vaccines (pneumococcal and tuberculosis).\\nRisk of death and of combined COVID-19 complications in all the vaccinated population and excluding people living in long-term care facilities (LTCF).\",\n \"[SUBTITLE] Principal Findings [SUBSECTION] We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\\nWe analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\\n[SUBTITLE] Comparison With Prior Work [SUBSECTION] Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\\nSome researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\\n[SUBTITLE] Limitations [SUBSECTION] We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\\nWe need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\\n[SUBTITLE] Conclusions [SUBSECTION] In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.\\nIn conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.\",\n \"We analyzed the negative outcomes among people with COVID-19 (N=309,039) and compared those who had received the influenza vaccine with those who were never vaccinated. Those who received the vaccine any time before having COVID-19 were at a higher risk of complications than those who were unvaccinated. We obtained similar results for those who were recently vaccinated (2019-2020 campaign) and for those who were systematically vaccinated (at least 3 years), and the same comparisons were carried out after excluding individuals living in LTCFs. We also obtained similar results on matching vaccinated and unvaccinated individuals. Thus, we did not find a possible link between receiving the influenza vaccine and presenting better clinical outcomes after a COVID-19 infection.\",\n \"Some researchers have studied this possible association. Massoudi and Mohit [28] conducted a study in a hospital in Iran including health care workers, with 80 of them COVID-19 cases confirmed through PCR or on the basis of their symptoms, and 181 of them were controls. They concluded that individuals who were confirmed cases were less likely to have received the 2019 influenza vaccine (OR 0.04, 95% CI 0.01-0.14), suggesting a protective association between the influenza vaccine and COVID-19. Their study had several limitations, such as the lack of availability of COVID-19 test kits or the samples limited to the workers of a single hospital [28].\\nCandelli et al [29] assessed 602 patients with COVID-19 enrolled at the emergency department in a hospital in Italy, of whom 24.9% had been previously vaccinated against influenza. They found that influenza immunization was independently associated with a lower risk of death at 60 days (OR 0.20, 95% CI 0.08-0.51), but not with a reduced need of endotracheal intubation (OR 0.73, 95% CI 0.35-1.56) [29].\\nA study conducted in Brazil [30] included 92,664 confirmed cases of COVID-19, of whom 31.1% had been recently vaccinated against influenza. They found that the vaccinated individuals were at a lower risk of needing intensive care for COVID-19 (OR 0.92, 95% CI 0.86-0.99), a lower risk of needing respiratory support (OR 0.81, 95% CI 0.74-0.88), and lower odds of mortality (OR 0.82, 95% CI 0.75-0.89) [30].\\nIn a systematic review [31] including 12 studies, the authors examined whether influenza vaccination affects the risk of being infected with SARS-CoV-2 and the risk of complicated illness or poor outcomes among patients with COVID-19, all of whom having been confirmed cases through PCR testing. They concluded that influenza vaccination is unlikely to be associated with an increase in the risk of COVID-19 infection or severity and the risk of associated death [31].\\nThere are reports from some countries with high influenza vaccination rates and high incidences of COVID-19 and mortality [32,33]. For instance, Kline et al [33] compared people vaccinated against influenza with unvaccinated individuals admitted to hospital for COVID-19, and they found no differences in the rate of admission to the intensive care unit, intubation, or other complications [33]. Our results follow these same trends in a cohort of the general population attended to in the PHC system and not only hospitalized patients.\",\n \"We need to consider that our results correspond to the first wave of the COVID-19 pandemic, when there were more negative outcomes and mortalities due to COVID-19 than in the subsequent waves in our setting; thus, this higher statistical power allowed us to detect differences. Furthermore, in subsequent waves, more confounders might have been present, such as COVID-19 vaccination or effects of the different SARS-CoV-2 variants, making it more difficult to manage their potential effect in the analysis of the outcomes of the infection.\\nWe also need to bear in mind that the target population for the influenza vaccine in our country are people older than 60 years, individuals with chronic comorbidities or immunodeficiency, and health care workers among others [34], some of them being at a high risk of COVID-19 complications, which is why confounding variables were used to adjust the logistic regression model [35]. Nevertheless, estimates of the effectiveness of the influenza vaccine have been frequently confounded, indicating that a different approach should be used with alternative study designs, different from the typical methods used to study drug exposure [36-38].\\nAmong other limitations of our study is the reliability of the COVID-19 diagnoses; we included individuals without a confirmed result, as during the first wave of the pandemic in our setting, PCR tests were not always performed. This limitation has been described in other studies including those conducted at the beginning of the pandemic when diagnostic tests for COVID-19 were not widely available and clinical algorithms were used to assess COVID-19 diagnoses [39]. We compared our number of COVID-19 cases with the official COVID-19 case numbers provided by the AQuAS during the pandemic [24]. Another limitation is the lack of hospital information: we could not capture ICU admissions, ventilation, or treatments administered upon admission, which clearly have an influence on the prognosis and outcomes of COVID-19. Finally, we have not conducted any subgroup analysis that could have indicated any condition potentially resulting in any benefit or harm from influenza vaccination.\",\n \"In conclusion, we were not able to establish a protective role of the immunity conferred by the influenza vaccine on the outcomes of COVID-19 infection. Nonetheless, our study adds more evidence to the analysis of the possible link between the quality of the conferred immunity and outcomes of COVID-19 infection, and it has some strengths, such as the large cohort size, its representativeness with respect to the general population, and the completeness of its sociodemographic data. We have already highlighted that our cohort comprises individuals who received care from the PHC system; hence, we have estimated the risk of complications for a different population from the hospitalized individuals who are usually assessed in multiple studies.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction","methods",null,null,null,null,null,null,null,"results","discussion",null,null,null,null],"string":"[\n \"introduction\",\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n \"discussion\",\n null,\n null,\n null,\n null\n]"},"keywords":{"kind":"list like","value":["SARS-CoV-2","COVID-19","influenza vaccines","pneumonia","electronic health records","primary health care","vaccination","public health","cohort study","epidemiology","eHeatlh","health outcome","mortality"],"string":"[\n \"SARS-CoV-2\",\n \"COVID-19\",\n \"influenza vaccines\",\n \"pneumonia\",\n \"electronic health records\",\n \"primary health care\",\n \"vaccination\",\n \"public health\",\n \"cohort study\",\n \"epidemiology\",\n \"eHeatlh\",\n \"health outcome\",\n \"mortality\"\n]"}}},{"rowIdx":369,"cells":{"title":{"kind":"string","value":"A RASSF8-AS1 based exosomal lncRNAs panel used for diagnostic and prognostic biomarkers for esophageal squamous cell carcinoma."},"pmid":{"kind":"string","value":"36266257"},"background_abstract":{"kind":"string","value":"Exosomal long non-coding RNA (lncRNA) has been shown to be potential biomarker for cancer diagnosis and follow up. However, little is known about its application in esophageal squamous cell carcinoma (ESCC) detection. Here, we sought to develop a novel diagnostic model based on serum exosomal lncRNAs to improve ESCC screening efficiency."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"A multiphase, case-control study was conducted among 140 ESCC patients and 140 healthy controls. Microarray screening was performed to acquire differentially expressed exosomal lncRNAs in the discovery phase. The diagnostic model Index I was constructed based on a panel of three lncRNAs using logistic regression in the training phase, and were confirmed in a subsequent validation phase. A receiver operating characteristic (ROC) curve was generated to calculate the diagnostic value. The effects of the selected lncRNAs level on ESCC mortality were evaluated using a Cox hazard regression model and Kaplan-Meier curve analysis, and the expression level with clinicopathological features was also calculated. Finally, we explored the oncogenic potential of candidate lncRNA RASSF8-AS1 in vitro and by target mRNA sequencing."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Index I was able to discriminate ESCC patients from healthy controls, and showed superiority to classic tumor biomarkers. Moreover, serum levels of the exosomal lncRNAs correlated with clinicopathological features and prognosis. The in vitro assays showed that RASSF8-AS1 played an oncogenic role in ESCC. Target mRNA scanning results suggested involvement of RASSF8-AS1 in tumor immunity and metabolism."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The newly identified serum exosomal lncRNAs could be used as new biomarkers for ESCC, and showed oncogenic potential in ESCC."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Esophageal Squamous Cell Carcinoma","RNA, Long Noncoding","Esophageal Neoplasms","Prognosis","Case-Control Studies","Gene Expression Regulation, Neoplastic","Biomarkers, Tumor","RNA, Messenger","Tumor Suppressor Proteins"],"string":"[\n \"Humans\",\n \"Esophageal Squamous Cell Carcinoma\",\n \"RNA, Long Noncoding\",\n \"Esophageal Neoplasms\",\n \"Prognosis\",\n \"Case-Control Studies\",\n \"Gene Expression Regulation, Neoplastic\",\n \"Biomarkers, Tumor\",\n \"RNA, Messenger\",\n \"Tumor Suppressor Proteins\"\n]"},"pmcid":{"kind":"string","value":"9715784"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Patients and samples [SUBSECTION] From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\nFrom 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\n[SUBTITLE] Exosome extraction and identification [SUBSECTION] A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\nA total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\n[SUBTITLE] \nRNA extraction and quantitative real‐time PCR (qRT‐PCR) [SUBSECTION] RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\nRNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\n[SUBTITLE] Microarray screening and analysis [SUBSECTION] Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\nMicroarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\n[SUBTITLE] Cell culture [SUBSECTION] Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\nEsophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\n[SUBTITLE] Stability analysis for exosomal lncRNAs panel [SUBSECTION] Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\nExosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\n[SUBTITLE] Transfection and RNA interfering [SUBSECTION] Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\nCells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\n[SUBTITLE] Cell proliferation, migration, and invasion assays [SUBSECTION] Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\nTransfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\n[SUBTITLE] \nRNA sequencing and enrichment analysis [SUBSECTION] Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\nCell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\n[SUBTITLE] Statistical analysis [SUBSECTION] All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\nAll data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Identification of exosomes in serum [SUBSECTION] A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\nA total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\n[SUBTITLE] Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls [SUBSECTION] The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\nThe overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\n[SUBTITLE] Evaluation of stability of exosomal lncRNA in serum samples under different conditions [SUBSECTION] Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\nGiven that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\n[SUBTITLE] Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I [SUBSECTION] To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\nTo further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\n[SUBTITLE] Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features [SUBSECTION] We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nWe further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\n[SUBTITLE] Prognostic significance of three serum lncRNAs panel [SUBSECTION] We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\nWe then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\n[SUBTITLE] In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning [SUBSECTION] As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\nAs lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Patients and samples","Exosome extraction and identification","\nRNA extraction and quantitative real‐time PCR (qRT‐PCR)","Microarray screening and analysis","Cell culture","Stability analysis for exosomal lncRNAs panel","Transfection and RNA interfering","Cell proliferation, migration, and invasion assays","\nRNA sequencing and enrichment analysis","Statistical analysis","Identification of exosomes in serum","Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls","Evaluation of stability of exosomal lncRNA in serum samples under different conditions","Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I","Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features","Prognostic significance of three serum lncRNAs panel","In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning","FUNDING INFORMATION"],"string":"[\n \"INTRODUCTION\",\n \"Patients and samples\",\n \"Exosome extraction and identification\",\n \"\\nRNA extraction and quantitative real‐time PCR (qRT‐PCR)\",\n \"Microarray screening and analysis\",\n \"Cell culture\",\n \"Stability analysis for exosomal lncRNAs panel\",\n \"Transfection and RNA interfering\",\n \"Cell proliferation, migration, and invasion assays\",\n \"\\nRNA sequencing and enrichment analysis\",\n \"Statistical analysis\",\n \"Identification of exosomes in serum\",\n \"Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls\",\n \"Evaluation of stability of exosomal lncRNA in serum samples under different conditions\",\n \"Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I\",\n \"Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features\",\n \"Prognostic significance of three serum lncRNAs panel\",\n \"In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning\",\n \"FUNDING INFORMATION\"\n]"},"other_sections_texts":{"kind":"list like","value":["Esophageal cancer is the eighth most common cancer and the sixth leading cause of malignancy deaths worldwide.\n1\n Squamous cell carcinoma is the main pathology type in China. The current standard for esophageal squamous cell carcinoma (ESCC) diagnosis depends on endoscopy and pathological biopsy.\n2\n However, the operation is relatively invasive and uncomfortable, resulting in delay or missing the best chance for most ESCC diagnosis and therapy. Liquid biopsy provides an alternative way of ESCC screening, but the traditional biomarkers used in the diagnosis of ESCC, such as serum squamous cell carcinoma antigen (SCC), cancer antigen 125 (CA125), carcinoembryonic antigen (CEA), and cytokeratin‐19‐fragment (CYFRA21‐1),\n3\n, \n4\n lack enough sensitivity and specificity, which limits their clinical application. For example, SCC may be also applied in other types of squamous cell carcinoma, such as lung, head and neck, and cervical squamous cell carcinoma. CEA is mainly used for the detection of colorectal cancer, while CYFRA21‐1 is usually used as a biomarker for diverse subtypes of lung and bladder cancer.\nExosomes are small (50–150 nm) membrane vesicles inside the cell, which may fuse with the membrane and are then released into the circulation.\n5\n, \n6\n They are detected in various body fluids such as serum, plasma, urine and saliva.\n7\n, \n8\n Exosomes execute communication functions intercellularly by transporting various biomolecules (such as DNAs, RNAs and proteins),\n9\n, \n10\n and the contents are protected by a phospholipid bilayer, making them suitable for liquid biopsy. Long noncoding RNAs (LncRNAs) is one of the exosome cargos,\n11\n and previous studies considered that lncRNAs in exosomes may be more stable for exosome protection against RNase.\n12\n, \n13\n LncRNAs detected in body fluids have been shown to be ideal noninvasive biomarkers for cancer diagnosis and prognosis. For example, lncRNA HOTAIR and SBF2‐AS1 in serum exosomes have been implicated for the diagnosis and prognosis assessment of glioblastoma.\n14\n, \n15\n Similarly, Xu et al.\n16\n found in an independent cohort study that ENSG00000258332.1 and LINC00635 in serum had higher sensitivity and specificity for the diagnosis of hepatocellular carcinoma in combination with AFP. Recently, Tao et al.\n17\n found that the expression levels of TBILA and AGAP2‐AS1 in the serum of NSCLC patients were significantly increased, and the combination with serum tumor biomarkers could further improve the diagnostic accuracy of NSCLC patients. Compared with the increasing evidence for exosomal lncRNAs in other types of cancer, little is known about their roles in ESCC. Furthermore, it has previously been reported that a joint model based on the union of multiple biomarkers may improve the diagnostic power, which prompted our study design.\n18\n, \n19\n\n\nHere, we used a multistep study to investigate the possibility that novel serum exosomal lncRNAs act as potential biomarkers for ESCC detection. We selected candidate lncRNAs by integrating microarray results derived from serum and tissue samples separately, as a previous study reported that serum exosomal lncRNAs may be derived from tumor tissues. We deduced that the lncRNAs simultaneously overexpressed in serum and tissue samples may be more likely to meet the criteria. Finally, a diagnostic model based on three lncRNAs panel was established and the association of the novel panel with clinicopathological features or prognosis was also evaluated. Furthermore, we found involvement of candidate lncRNA RASSF8‐AS1 in ESCC malignant phenotypes, and the combined results provide new insights into ESCC diagnosis and target therapy.","From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.","A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).","RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.","Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.","Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.","Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.","Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.","Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.","Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.","All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.","A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.","The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .","Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.","To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.","We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.","We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.","As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.","This research was supported by the National Natural Science Foundation for Youth of China (no. 81702444), China Postdoctoral Science Foundation 2018M643884 and 2020T130128ZX."],"string":"[\n \"Esophageal cancer is the eighth most common cancer and the sixth leading cause of malignancy deaths worldwide.\\n1\\n Squamous cell carcinoma is the main pathology type in China. The current standard for esophageal squamous cell carcinoma (ESCC) diagnosis depends on endoscopy and pathological biopsy.\\n2\\n However, the operation is relatively invasive and uncomfortable, resulting in delay or missing the best chance for most ESCC diagnosis and therapy. Liquid biopsy provides an alternative way of ESCC screening, but the traditional biomarkers used in the diagnosis of ESCC, such as serum squamous cell carcinoma antigen (SCC), cancer antigen 125 (CA125), carcinoembryonic antigen (CEA), and cytokeratin‐19‐fragment (CYFRA21‐1),\\n3\\n, \\n4\\n lack enough sensitivity and specificity, which limits their clinical application. For example, SCC may be also applied in other types of squamous cell carcinoma, such as lung, head and neck, and cervical squamous cell carcinoma. CEA is mainly used for the detection of colorectal cancer, while CYFRA21‐1 is usually used as a biomarker for diverse subtypes of lung and bladder cancer.\\nExosomes are small (50–150 nm) membrane vesicles inside the cell, which may fuse with the membrane and are then released into the circulation.\\n5\\n, \\n6\\n They are detected in various body fluids such as serum, plasma, urine and saliva.\\n7\\n, \\n8\\n Exosomes execute communication functions intercellularly by transporting various biomolecules (such as DNAs, RNAs and proteins),\\n9\\n, \\n10\\n and the contents are protected by a phospholipid bilayer, making them suitable for liquid biopsy. Long noncoding RNAs (LncRNAs) is one of the exosome cargos,\\n11\\n and previous studies considered that lncRNAs in exosomes may be more stable for exosome protection against RNase.\\n12\\n, \\n13\\n LncRNAs detected in body fluids have been shown to be ideal noninvasive biomarkers for cancer diagnosis and prognosis. For example, lncRNA HOTAIR and SBF2‐AS1 in serum exosomes have been implicated for the diagnosis and prognosis assessment of glioblastoma.\\n14\\n, \\n15\\n Similarly, Xu et al.\\n16\\n found in an independent cohort study that ENSG00000258332.1 and LINC00635 in serum had higher sensitivity and specificity for the diagnosis of hepatocellular carcinoma in combination with AFP. Recently, Tao et al.\\n17\\n found that the expression levels of TBILA and AGAP2‐AS1 in the serum of NSCLC patients were significantly increased, and the combination with serum tumor biomarkers could further improve the diagnostic accuracy of NSCLC patients. Compared with the increasing evidence for exosomal lncRNAs in other types of cancer, little is known about their roles in ESCC. Furthermore, it has previously been reported that a joint model based on the union of multiple biomarkers may improve the diagnostic power, which prompted our study design.\\n18\\n, \\n19\\n\\n\\nHere, we used a multistep study to investigate the possibility that novel serum exosomal lncRNAs act as potential biomarkers for ESCC detection. We selected candidate lncRNAs by integrating microarray results derived from serum and tissue samples separately, as a previous study reported that serum exosomal lncRNAs may be derived from tumor tissues. We deduced that the lncRNAs simultaneously overexpressed in serum and tissue samples may be more likely to meet the criteria. Finally, a diagnostic model based on three lncRNAs panel was established and the association of the novel panel with clinicopathological features or prognosis was also evaluated. Furthermore, we found involvement of candidate lncRNA RASSF8‐AS1 in ESCC malignant phenotypes, and the combined results provide new insights into ESCC diagnosis and target therapy.\",\n \"From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\",\n \"A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\",\n \"RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\\n20\\n, \\n21\\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\",\n \"Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\",\n \"Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\",\n \"Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\",\n \"Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\",\n \"Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\",\n \"Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\",\n \"All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\",\n \"A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\",\n \"The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\",\n \"Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\",\n \"To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\\n22\\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\\n23\\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\",\n \"We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\",\n \"We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\",\n \"As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\",\n \"This research was supported by the National Natural Science Foundation for Youth of China (no. 81702444), China Postdoctoral Science Foundation 2018M643884 and 2020T130128ZX.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Patients and samples","Exosome extraction and identification","\nRNA extraction and quantitative real‐time PCR (qRT‐PCR)","Microarray screening and analysis","Cell culture","Stability analysis for exosomal lncRNAs panel","Transfection and RNA interfering","Cell proliferation, migration, and invasion assays","\nRNA sequencing and enrichment analysis","Statistical analysis","RESULTS","Identification of exosomes in serum","Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls","Evaluation of stability of exosomal lncRNA in serum samples under different conditions","Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I","Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features","Prognostic significance of three serum lncRNAs panel","In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning","DISCUSSION","FUNDING INFORMATION","CONFLICT OF INTEREST","Supporting information"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Patients and samples\",\n \"Exosome extraction and identification\",\n \"\\nRNA extraction and quantitative real‐time PCR (qRT‐PCR)\",\n \"Microarray screening and analysis\",\n \"Cell culture\",\n \"Stability analysis for exosomal lncRNAs panel\",\n \"Transfection and RNA interfering\",\n \"Cell proliferation, migration, and invasion assays\",\n \"\\nRNA sequencing and enrichment analysis\",\n \"Statistical analysis\",\n \"RESULTS\",\n \"Identification of exosomes in serum\",\n \"Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls\",\n \"Evaluation of stability of exosomal lncRNA in serum samples under different conditions\",\n \"Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I\",\n \"Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features\",\n \"Prognostic significance of three serum lncRNAs panel\",\n \"In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning\",\n \"DISCUSSION\",\n \"FUNDING INFORMATION\",\n \"CONFLICT OF INTEREST\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Esophageal cancer is the eighth most common cancer and the sixth leading cause of malignancy deaths worldwide.\n1\n Squamous cell carcinoma is the main pathology type in China. The current standard for esophageal squamous cell carcinoma (ESCC) diagnosis depends on endoscopy and pathological biopsy.\n2\n However, the operation is relatively invasive and uncomfortable, resulting in delay or missing the best chance for most ESCC diagnosis and therapy. Liquid biopsy provides an alternative way of ESCC screening, but the traditional biomarkers used in the diagnosis of ESCC, such as serum squamous cell carcinoma antigen (SCC), cancer antigen 125 (CA125), carcinoembryonic antigen (CEA), and cytokeratin‐19‐fragment (CYFRA21‐1),\n3\n, \n4\n lack enough sensitivity and specificity, which limits their clinical application. For example, SCC may be also applied in other types of squamous cell carcinoma, such as lung, head and neck, and cervical squamous cell carcinoma. CEA is mainly used for the detection of colorectal cancer, while CYFRA21‐1 is usually used as a biomarker for diverse subtypes of lung and bladder cancer.\nExosomes are small (50–150 nm) membrane vesicles inside the cell, which may fuse with the membrane and are then released into the circulation.\n5\n, \n6\n They are detected in various body fluids such as serum, plasma, urine and saliva.\n7\n, \n8\n Exosomes execute communication functions intercellularly by transporting various biomolecules (such as DNAs, RNAs and proteins),\n9\n, \n10\n and the contents are protected by a phospholipid bilayer, making them suitable for liquid biopsy. Long noncoding RNAs (LncRNAs) is one of the exosome cargos,\n11\n and previous studies considered that lncRNAs in exosomes may be more stable for exosome protection against RNase.\n12\n, \n13\n LncRNAs detected in body fluids have been shown to be ideal noninvasive biomarkers for cancer diagnosis and prognosis. For example, lncRNA HOTAIR and SBF2‐AS1 in serum exosomes have been implicated for the diagnosis and prognosis assessment of glioblastoma.\n14\n, \n15\n Similarly, Xu et al.\n16\n found in an independent cohort study that ENSG00000258332.1 and LINC00635 in serum had higher sensitivity and specificity for the diagnosis of hepatocellular carcinoma in combination with AFP. Recently, Tao et al.\n17\n found that the expression levels of TBILA and AGAP2‐AS1 in the serum of NSCLC patients were significantly increased, and the combination with serum tumor biomarkers could further improve the diagnostic accuracy of NSCLC patients. Compared with the increasing evidence for exosomal lncRNAs in other types of cancer, little is known about their roles in ESCC. Furthermore, it has previously been reported that a joint model based on the union of multiple biomarkers may improve the diagnostic power, which prompted our study design.\n18\n, \n19\n\n\nHere, we used a multistep study to investigate the possibility that novel serum exosomal lncRNAs act as potential biomarkers for ESCC detection. We selected candidate lncRNAs by integrating microarray results derived from serum and tissue samples separately, as a previous study reported that serum exosomal lncRNAs may be derived from tumor tissues. We deduced that the lncRNAs simultaneously overexpressed in serum and tissue samples may be more likely to meet the criteria. Finally, a diagnostic model based on three lncRNAs panel was established and the association of the novel panel with clinicopathological features or prognosis was also evaluated. Furthermore, we found involvement of candidate lncRNA RASSF8‐AS1 in ESCC malignant phenotypes, and the combined results provide new insights into ESCC diagnosis and target therapy.","[SUBTITLE] Patients and samples [SUBSECTION] From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\nFrom 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\n[SUBTITLE] Exosome extraction and identification [SUBSECTION] A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\nA total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\n[SUBTITLE] \nRNA extraction and quantitative real‐time PCR (qRT‐PCR) [SUBSECTION] RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\nRNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\n[SUBTITLE] Microarray screening and analysis [SUBSECTION] Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\nMicroarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\n[SUBTITLE] Cell culture [SUBSECTION] Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\nEsophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\n[SUBTITLE] Stability analysis for exosomal lncRNAs panel [SUBSECTION] Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\nExosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\n[SUBTITLE] Transfection and RNA interfering [SUBSECTION] Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\nCells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\n[SUBTITLE] Cell proliferation, migration, and invasion assays [SUBSECTION] Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\nTransfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\n[SUBTITLE] \nRNA sequencing and enrichment analysis [SUBSECTION] Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\nCell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\n[SUBTITLE] Statistical analysis [SUBSECTION] All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\nAll data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.","From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.","A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).","RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\n20\n, \n21\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.","Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.","Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.","Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.","Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.","Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.","Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.","All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.","[SUBTITLE] Identification of exosomes in serum [SUBSECTION] A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\nA total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\n[SUBTITLE] Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls [SUBSECTION] The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\nThe overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\n[SUBTITLE] Evaluation of stability of exosomal lncRNA in serum samples under different conditions [SUBSECTION] Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\nGiven that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\n[SUBTITLE] Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I [SUBSECTION] To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\nTo further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\n[SUBTITLE] Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features [SUBSECTION] We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nWe further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\n[SUBTITLE] Prognostic significance of three serum lncRNAs panel [SUBSECTION] We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\nWe then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\n[SUBTITLE] In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning [SUBSECTION] As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\nAs lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.","A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.","The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .","Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.","To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\n22\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\n23\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.","We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.","We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.","As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\nBold values means p‐value < 0.05.\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.","Delayed diagnosis is the main cause for poor prognosis and rapid progression of ESCC. The five‐year survival rate for patients with early diagnosis of ESCC is >90%, while survival rate with late detection is <20%.\n24\n Endoscopic surveillance and biopsy examination are current applied approaches for ESCC diagnosis and monitoring, but are not popular because of the high cost and level of discomfort, while imaging technology is limited for the low sensitivity, thereby highlighting the urgent need to develop new, relatively noninvasive and reliable tools to improve the early diagnosis of ESCC.\nThere are currently no widely accepted serum biomarkers for ESCC screening, surveillance for treatment response and recurrence. Traditional tumor biomarkers are inadequate in ESCC detection as they have poor sensitivity and specificity.\nIn recent years, the emergence of novel liquid tumor markers has provided a new solution for the early screening of tumors. The discovery of circulating tumor cells (CTC),\n25\n circulating tumor cell DNA (ctDNA),\n26\n circulating cell‐free DNA (cfDNA),\n27\n and exosomes can further improve the efficacy of noninvasive screening. These humoral markers reflect the molecular and genetic information, as well as temporal and spatial heterogeneity of tumors, making the dynamic monitoring of tumor status possible. Among the currently acceptable noninvasive markers, the total amount of CTC is relatively low, while the specificity of ctDNA and cfDNA is insufficient, making the clinical application full of challenges. With the development of exosome isolation and enrichment technologies, exosome diagnosis has become a rising star in the field of liquid biopsy. Benefiting from the protection of the phospholipid bilayer on the surface of exosomes, the exosome contents have good stability, especially the RNAs can be protected by the membrane to avoid being degraded by RNase. The discovery of noncoding RNAs has enriched the understanding of tumorigenesis mechanisms. The role of exosome‐transported noncoding RNAs in tumor metastasis, drug resistance, and radio‐resistance across cells has received increasing attention. Studies have shown that the number of LncRNAs is huge, dozens of times that of currently known mRNAs, with high tissue specificity, which provides ideal conditions for the development of noninvasive screening. It has been reported that serum exosomal biomarkers may be derived from tumor tissues,\n28\n therefore in this study we also incorporated the upregulated lncRNAs from tissue microarray. The linear association between serum and tissue samples for the candidate lncRNAs confirmed the assumption. Finally, we identified a novel panel of exosomal lncRNAs for ESCC liquid biopsy through a multiphase study.\nThe lncRNAs‐based composite Index I showed significantly higher diagnostic power than traditional tumor biomarker based composite Index II in two independent cohorts. Furthermore, the novel lncRNAs signature was stably detectable in serum, supporting the clinically applicable value. We also noted serum lncRNAs declined post‐surgery, indicating that Index I may reflect tumor burden of ESCC patients, which should be used as a follow‐up parameter for ESCC.\nAmong the newly identified lncRNA panel, RASSF8‐AS1 showed the best diagnostic power for all ESCC patients, suggesting its oncogenic role in ESCC. Currently, the function of RASSF8‐AS1 has only been reported in laryngeal squamous cell carcinoma.\n29\n Our results showed its oncogenic role in ESCC for the first time and revealed the underlying mechanisms by target mRNA scanning. Similarly, the role of LIN00958 in ESCC has been recently reported, which also verified the oncogenic role of LIN00958, suggesting our results are reasonable.\n30\n\n\nWe also note the limitations of this study. First, we included only a relatively small number of patients and healthy controls, and the ESCC patients were diagnosed with cancer as they had already been diagnosed pathologically earlier, which should be verified in a much larger sample size. Second, we conducted a retrospective, single‐center, cross‐sectional study, which may have introduced unavoidable selection bias. A prospective multicenter study is needed to further confirm the diagnostic efficiency of Index I, and to check the superiority to traditional biomarkers. Next, how lncRNA enters the exosome, how lncRNA encapsulated exosomes play a role in tumor progress and distant migration call for more attention. In recent years, the role of exosomes in mediating distant tumor metastasis has gained emphasis,\n31\n, \n32\n and our study revealed high expression of serum lncRNA RASSF8‐AS1 was also correlated with metastasis, and our sequencing results suggested its involvement in immune response and cell metabolism. How our newly discovered exosomal lncRNAs play a role in distant metastasis of ESCC deserves further investigation.","This research was supported by the National Natural Science Foundation for Youth of China (no. 81702444), China Postdoctoral Science Foundation 2018M643884 and 2020T130128ZX.","All authors declare no conflict of interest.","\nTable S1 The sequences of qRT‐PCR primers.\n\nTable S2 The sequences of siRNAs.\n\nTable S3 Demographic and clinicopathologic characteristics of the study subjects.\n\nTable S4 Results of ROC curves for individual biomarkers.\nClick here for additional data file.\n\nFigure S1 Characterization of exosomes isolated from serum samples. (A) TEM images of exosomes. HV = 100.0 kV, Direct Mag: 100000× (scale bars 100 nm). Exosomes were winkled oval or spherical in shape under TEM. (B) Size and concentration distribution of exosomes were analyzed by NTA.\nClick here for additional data file.\n\nFigure S2 Bioinformatic analysis of differential mRNAs. (A) Hierarchical clustering of the high‐throughput microarray performed for serum mRNA. (B–D) Gene ontology enrichments of biological process (BP), Cellular component (CC) and molecular function (MF) for differential mRNAs.\nClick here for additional data file.\n\nFigure S3 Big data mining for the candidate lncRNAs. (A–C) Expression of serum exosomal lncRNAs in pan‐cancer compared with healthy donors. (D) Tissues with higher stages exhibit roughly gradually elevated expression of RASSF8‐AS1. *p < 0.05, **p < 0.01, ***p < 0.001.\nClick here for additional data file.\n\nFigure S4 RNA‐sequencing results of RSSF8‐AS1 targets. (A) Volcano plots for the differential mRNAs. (B) Hierarchical Clustering for the differentially expressed RNA. (C, D) Gene Oncology and pathway enrichment analysis for the positively regulated mRNAs of RASSF8‐AS1. (E, F) Oncology and pathway enrichment analysis for the negatively regulated mRNAs of RASSF8‐AS1.\nClick here for additional data file."],"string":"[\n \"Esophageal cancer is the eighth most common cancer and the sixth leading cause of malignancy deaths worldwide.\\n1\\n Squamous cell carcinoma is the main pathology type in China. The current standard for esophageal squamous cell carcinoma (ESCC) diagnosis depends on endoscopy and pathological biopsy.\\n2\\n However, the operation is relatively invasive and uncomfortable, resulting in delay or missing the best chance for most ESCC diagnosis and therapy. Liquid biopsy provides an alternative way of ESCC screening, but the traditional biomarkers used in the diagnosis of ESCC, such as serum squamous cell carcinoma antigen (SCC), cancer antigen 125 (CA125), carcinoembryonic antigen (CEA), and cytokeratin‐19‐fragment (CYFRA21‐1),\\n3\\n, \\n4\\n lack enough sensitivity and specificity, which limits their clinical application. For example, SCC may be also applied in other types of squamous cell carcinoma, such as lung, head and neck, and cervical squamous cell carcinoma. CEA is mainly used for the detection of colorectal cancer, while CYFRA21‐1 is usually used as a biomarker for diverse subtypes of lung and bladder cancer.\\nExosomes are small (50–150 nm) membrane vesicles inside the cell, which may fuse with the membrane and are then released into the circulation.\\n5\\n, \\n6\\n They are detected in various body fluids such as serum, plasma, urine and saliva.\\n7\\n, \\n8\\n Exosomes execute communication functions intercellularly by transporting various biomolecules (such as DNAs, RNAs and proteins),\\n9\\n, \\n10\\n and the contents are protected by a phospholipid bilayer, making them suitable for liquid biopsy. Long noncoding RNAs (LncRNAs) is one of the exosome cargos,\\n11\\n and previous studies considered that lncRNAs in exosomes may be more stable for exosome protection against RNase.\\n12\\n, \\n13\\n LncRNAs detected in body fluids have been shown to be ideal noninvasive biomarkers for cancer diagnosis and prognosis. For example, lncRNA HOTAIR and SBF2‐AS1 in serum exosomes have been implicated for the diagnosis and prognosis assessment of glioblastoma.\\n14\\n, \\n15\\n Similarly, Xu et al.\\n16\\n found in an independent cohort study that ENSG00000258332.1 and LINC00635 in serum had higher sensitivity and specificity for the diagnosis of hepatocellular carcinoma in combination with AFP. Recently, Tao et al.\\n17\\n found that the expression levels of TBILA and AGAP2‐AS1 in the serum of NSCLC patients were significantly increased, and the combination with serum tumor biomarkers could further improve the diagnostic accuracy of NSCLC patients. Compared with the increasing evidence for exosomal lncRNAs in other types of cancer, little is known about their roles in ESCC. Furthermore, it has previously been reported that a joint model based on the union of multiple biomarkers may improve the diagnostic power, which prompted our study design.\\n18\\n, \\n19\\n\\n\\nHere, we used a multistep study to investigate the possibility that novel serum exosomal lncRNAs act as potential biomarkers for ESCC detection. We selected candidate lncRNAs by integrating microarray results derived from serum and tissue samples separately, as a previous study reported that serum exosomal lncRNAs may be derived from tumor tissues. We deduced that the lncRNAs simultaneously overexpressed in serum and tissue samples may be more likely to meet the criteria. Finally, a diagnostic model based on three lncRNAs panel was established and the association of the novel panel with clinicopathological features or prognosis was also evaluated. Furthermore, we found involvement of candidate lncRNA RASSF8‐AS1 in ESCC malignant phenotypes, and the combined results provide new insights into ESCC diagnosis and target therapy.\",\n \"[SUBTITLE] Patients and samples [SUBSECTION] From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\\nFrom 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\\n[SUBTITLE] Exosome extraction and identification [SUBSECTION] A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\\nA total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\\n[SUBTITLE] \\nRNA extraction and quantitative real‐time PCR (qRT‐PCR) [SUBSECTION] RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\\n20\\n, \\n21\\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\\nRNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\\n20\\n, \\n21\\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\\n[SUBTITLE] Microarray screening and analysis [SUBSECTION] Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\\nMicroarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\\n[SUBTITLE] Cell culture [SUBSECTION] Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\\nEsophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\\n[SUBTITLE] Stability analysis for exosomal lncRNAs panel [SUBSECTION] Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\\nExosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\\n[SUBTITLE] Transfection and RNA interfering [SUBSECTION] Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\\nCells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\\n[SUBTITLE] Cell proliferation, migration, and invasion assays [SUBSECTION] Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\\nTransfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\\n[SUBTITLE] \\nRNA sequencing and enrichment analysis [SUBSECTION] Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\\nCell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\\n[SUBTITLE] Statistical analysis [SUBSECTION] All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\\nAll data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\",\n \"From 2017 to 2021, 140 ESCC patients and 140 healthy individuals were randomly recruited from Jinling Hospital of Nanjing Medical University, each of whom provided 5 ml of whole blood at indicated timepoints. Neither patient received preoperative chemo−/radiotherapy, and the overall survival (OS) was measured from the date of diagnosis to the date of death due to any cause or the date of the most recent follow‐up. All patients were followed every 3 months for the first year, every 6 months for the next 2 years, and then annually. The end of the follow‐up was November 31, 2021. The study was approved by the ethics committee of Jinling hospital, and all enrolled participants signed informed consent. All serum samples were stored at −80°C until use for total RNA extraction. The pathological stage was determined according to the eighth edition of the Union for International Cancer Control TNM staging system.\",\n \"A total of 1.5 ml of serum was centrifuged at 2000 × g, 4°C for 30 min, and the supernatant was then recentrifuged at 10 000 × g, 4°C for 45 min. The supernatant was then collected and filtered through a membrane of 0.22 μm pore size. The filtrate was collected and further subjected to ultracentrifugation at 10 000 × g, 4°C for 70 min. The precipitate was resuspended in 10 ml of precooling 1 × PBS, and ultracentrifugation repeated. Finally, the precipitate was resuspended in 1.5 ml of precooling 1 × PBS. The isolated exosomes were characterized using transmission electron microscopy (TEM). Exosome concentration‐particle size testing (nanoparticle tracking analysis [NTA]) was performed by Zeta View (Particle Metrix Ltd.).\",\n \"RNA was extracted from exosomes or tissues using an exosome RNA isolation kit according to protocols (Invitrogen). The extracted RNA was synthesized into cDNA by the PrimeScript RT Master Mix (Takara). RT‐qPCR was performed to detect the relative RNA levels of target genes. β‐actin was used as an internal reference gene for analysis. The relative expression levels of the lncRNAs were calculated using 2−∆∆Ct.\\n20\\n, \\n21\\n The sequences of primers used for qRT‐PCR of the lncRNAs are listed in Table S1.\",\n \"Microarray screening was performed with Arraystar LncRNA_8 × 60 k (Arraystar), which contains both lncRNA and mRNA probes. Sample labeling and array hybridization were performed according to the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technology). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX version 12.1 software package (Agilent Technologies). After quantile normalization of the raw data, LncRNAs and mRNAs that at least four out of eight samples have flags in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed LncRNAs and mRNAs with statistical significance between the two groups were identified through p‐value/FDR filtering.\",\n \"Esophageal squamous cell carcinoma cell lines (TE1, TE13, KYSE150, KYSE140) and normal human esophageal epithelial cells (HET‐1A) were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China). These cells were cultured in 1640 medium (KeyGene) containing 10% fetal bovine serum (FBS) (KeyGene), at 37°C in a 5% CO2 atmosphere.\",\n \"Exosome samples (n = 15) were allocated to aliquots for different processing, including storage under different temperature, treated with RNase A, and subjected to freeze–thaw cycles and acidic/alkaline environments, to test to the stability of exosomal lncRNAs, which is essential for clinical use.\",\n \"Cells were seeded on six‐well plates (2 × 105/well) with 1640 medium supplemented with 10% FBS overnight. Then, siRNA was transfected at a final concentration of 100 nM using lipofectamine 2000 (Invitrogen), and 24 h post transfection cells were harvested and subsequently RNA was extracted by Trizol reagent and the interfering efficiency was determined by RT‐qPCR. Sequences of siRNAs and negative control are provided in Table S2.\",\n \"Transfected cells were collected and grown in 96‐well plates (2 × 103 cells per well) overnight and cell proliferation was detected using cell counting kit‐8 (CCK‐8). The reaction products were measured at 450 nm according to the manufacturer's instructions. For migration/invasion assays, 5 × 104 cells were incubated using Transwell kits (8 mm PET, 24‐well Millicell) or matrix‐coated inserts (BD Biosciences). After 24 or 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotim). Imaging and counting of migrated cells were conducted by microscopy.\",\n \"Cell transfection and RNA interfering were performed as described above. Then, 24 h post transfection, cells were harvested and RNA was extracted using Trizol reagent (Invitrogen). RNA‐seq libraries were prepared using the Illumina RNA‐seq preparation kit and sequenced on illumina nova seq 6000 sequencer. The differentially expressed RNAs (fold change >1.5, p < 0.05) were retrieved and subjected to gene ontology and KEGG pathway analysis.\",\n \"All data were statistically analyzed with SPSS 26.0 (Chicago) and visualized with GraphPad Prism version 8.0 (GraphPad Software). Differences in lncRNAs expression between groups were determined using the Mann–Whitney unpaired t‐test or paired t‐test. The χ2 or Fisher's exact test was used to compare the clinical characteristics of the groups. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of selected biomarkers. The optimal cutoff point was determined using Youden's index. DeLong's test was used to evaluate the statistical significance when comparing differences of AUCs. Using the binary status of enrolled participants (patients or controls) as the dependent variable, the regression coefficient for each lncRNA was estimated by a univariate logistic regression model, and was used as the weighting to construct the diagnostic index. The diagnostic index developed in the training phase was also directly applied to the validation phase to evaluate the clinical utility of the identified lncRNAs individually or in combination. A forward conditional logistic regression model was used to analyze the association between selected biomarkers and ESCC mortality. The long‐term prognostic values of selected biomarkers were evaluated using Cox's hazard regression model and Kaplan–Meier curve analysis. Univariate analysis was used to explore clinicopathological factors (i.e., tumor depth, stage, size) associated with OS. Factors with p‐value <0.05 were selected for the multivariate analysis with Cox's proportional hazard regression model, hazard ratio and 95% confidence interval (CI) were calculated. All p‐values <0.05 were considered statistically significant.\",\n \"[SUBTITLE] Identification of exosomes in serum [SUBSECTION] A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\\nA total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\\n[SUBTITLE] Discovery phase: Genome‐wide profiling to identify differentially expressed lncRNAs between ESCC and healthy individual controls [SUBSECTION] The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\\nThe overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\\n[SUBTITLE] Evaluation of stability of exosomal lncRNA in serum samples under different conditions [SUBSECTION] Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\\nGiven that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\\n[SUBTITLE] Training phase: A panel of serum lncRNAs‐based diagnostic indicators Index I [SUBSECTION] To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\\n22\\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\\n23\\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\\nTo further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\\n22\\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\\n23\\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\\n[SUBTITLE] Validation phase: The lncRNAs based model can distinguish ESCC from healthy control and the levels of lncRNA panel correlated with clinicopathological features [SUBSECTION] We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\nWe further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\n[SUBTITLE] Prognostic significance of three serum lncRNAs panel [SUBSECTION] We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\\nWe then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\\n[SUBTITLE] In vitro analysis of the oncogenic potential of RASSF8‐AS1 and target scanning [SUBSECTION] As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\\nAs lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\",\n \"A total of 280 participants, including 140 ESCC patients and 140 healthy controls were enrolled in the study, as shown in Table S3. To verify whether the exosomes were successfully isolated from serum, we characterized the isolated exosomes using TEM and NTA. The results showed that the isolated exosomes had a bilayer membrane structure (Figure S1 A), and the particle size distribution of exosomes was about 30–150 nm in diameter (Figure S1 B), meeting the criteria of exosome patterns.\",\n \"The overall workflow of this study is illustrated in Figure 1a. In the discovery phase, we performed LncRNA microarray analysis on serum exosome samples from four preoperative ESCC patients and matched healthy controls. The top 100 differential lncRNAs identified in serum exosome samples from four preoperative ESCC patients and healthy control are shown in Figure 1b (FDR <0.005, Fold change >2, GEO accession number: GSE192662). Circos plots globally and genome‐widely displayed kinds of detected lncRNAs in ESCC and control serum samples (Figure 1c). A previous microarray comparing five pairs of tumor and paracancerous tissues (GSE89102) was also included for further analysis. To achieve more reliable aberrantly expressed lncRNAs, we adopted a more restrictive filtering criteria (FDR <0.005, fold change >5) for both datasets, and integrated the two datasets to finally obtain five (AC098818.2, RASSF8‐AS1, LINC00958, GMDS‐DT and AL591721.1) candidate genes with significant overexpression in ESCC serum samples and cancerous tissues (Figure 1d).\\nMicroarray screening results of differential lncRNAs and function prediction. (a) Experimental design overview. (b) Heatmap results of microarray analysis for four preoperative esophageal squamous cell carcinoma (ESCC) patients and matched normal healthy serum samples. (c) Circos plots of lncRNAs in the human genome (hg19). The outer tracks represent the cytoband ideogram of chromosome. For the two tracks, the outer one (blue) represents the levels of lncRNAs in normal tissues and the inner one (red) represents the levels in ESCC tissues. (d) Venn diagram showing consistent upregulation of five lncRNAs in both microarray analyses. (e, f) Signal pathway networks of mRNAs involved in lncRNAs‐mRNAs relationships.\\nTo investigate the potential role of the candidate lncRNAs in ESCC, we analyzed mRNA results from the same microarray chip derived from serum samples and performed pathway analysis (Figure S2 A–D). By lncRNAs‐mRNA coexpression network analysis, most of the candidate lncRNAs‐related mRNAs could be enriched in pathways such as MAPK, Ras, Wnt, metabolic pathways and HIF‐1α, which have been identified to be involved in ESCC carcinogenesis (Figure 1e,f).\\nTo validate the microarray data, we examined the expression of five candidate lncRNAs in a pilot cohort containing serum exosome samples (n = 20 for ESCC and healthy control separately), and 30 paired tissues, respectively. The qRT‐PCR results showed that the expression of AC098818.2, RASSF8‐AS1 and LINC00958 being significantly increased in ESCC patients' serum compared with healthy, while no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2a). Compared with normal esophageal tissues, AC098818.2, RASSF8‐AS1 and LINC00958 expression was significantly upregulated in tumor tissues, whereas the expression level of GMDS‐DT was significantly downregulated, and no significant difference was seen in the expression of AL591721.1 (Figure 2b). Moreover, we examined the expression levels of these five lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and normal epithelial cell HET‐1A. Compared with HET‐1A, the expression of AC098818.2, RASSF8‐AS1 and LINC00958 was upregulated in the ESCC cell lines, while there was no significant difference in the expression of AL591721.1 and GMDS‐DT (Figure 2c). Meanwhile, by big data mining through ExoRbase, we found that AC098818.2, RASSF8‐AS1 and LINC00958 were significantly overexpressed in ESCC blood exosomes compared with healthy donors and other types of cancers (Figure S3 A‐C), indicating excellent specificity in ESCC screening. By searching in the GEPIA database which obtained expression data from TCGA, a gradually increased trend of RASSF8‐AS1 expression toward TNM stages was found (Figure S3 D). Furthermore, we detected expression level of the three candidates in 12 pairs of serum samples and corresponding cancerous tissues, and found moderate correlation for each lncRNA (Figure 2d‐f).\\nIdentification of candidate five lncRNAs expression. (a) The expression levels of the five candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 20) and normal serum samples (n = 20). (b) The expression levels of the five candidate lncRNAs in paired tumors and adjacent normal tissues (n = 30). (c) The expression levels of the five candidate lncRNAs in ESCC cell lines (TE1, TE13, KYSE140 and KYSE150) and the human esophageal epithelial cells line (HET‐1A). (d–f) Spearman's correlation analysis between expression levels of lncRNAs in tumor tissues and paired serum samples of ESCC. *p < 0.05, **p < 0.01 .\",\n \"Given that stability is a key prerequisite for body fluid tumor biomarkers, we evaluated the stability of exosomal lncRNAs in serum samples under different conditions. Serum samples were subjected to extreme conditions, such as incubation at different temperatures for 3 h, repeated freeze–thaw cycles, RNase A digestion for 3 and 6 h, and exposure to acidic or alkaline environments. Interestingly, these treatments had little or no effect on the level of these serum lncRNAs (Figure 3), indicating the lncRNAs panel was valuable for clinical application.\\nStability of plasma lncRNAs under extreme conditions. (a) Incubation at different temperature for 3 h. (b) Repeated freeze–thaw cycles. (c) RNase A digestion for 3 and 6 h. (d) Exposure to acidic or alkaline environments.\",\n \"To further investigate the diagnostic ability of the three candidate lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) for ESCC, we determined the relative expression levels of these three lncRNAs in serum samples of 40 ESCC patients and 40 healthy individuals and performed ROC curve analysis to evaluate the diagnostic capability for each lncRNA alone. The results showed that the expression levels of these lncRNAs were significantly elevated in ESCC patients (Figure 4a), with AUC values ranging from 0.69 to 0.78 (Figure 4b). Table S4 summarizes in detail the AUC values and 95% CI of each individual lncRNA. The optimal threshold for each lncRNA was determined by the Youden's index.\\n22\\n Serum lncRNA RASSF8‐AS1 was the best diagnostic marker, with AUC values of 0.78 and 95% CI: 0.67–0.88, respectively.\\nTraining phase: A panel of serum lncRNAs‐based diagnostic indicators Index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) serum (n = 40) and normal serum samples (n = 40). (b) Receiver operating characteristic (ROC) curves to evaluate the sensitivity and specificity of three candidate lncRNAs to discriminate ESCC from healthy controls. (c) Paired comparison of the three candidate lncRNAs level before and after surgery (n = 20). (d) ROC curves to evaluate the sensitivity and specificity of traditional tumor markers to discriminate ESCC from healthy control. (e) Comparison of the diagnostic performance of Index I and Index II. *p < 0.05, **p < 0.01.\\nTo evaluate whether the expression levels of the lncRNAs panel could be used to monitor tumor dynamics, we collected paired serum samples from 20 pre‐ and postoperative patients. By paired comparison, we found that all three lncRNAs levels descended significantly after surgery (Figure 4c), reflecting potential utility in tumor burden monitoring during the therapy period.\\nMeanwhile, we explored the diagnostic value of several commonly used tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) in ESCC detection. The ROC curves showed that the AUC values of these markers ranged from 0.51 to 0.53 (Figure 4d), and Table S4 showed that CA199 had the best diagnostic ability with AUC values of 0.53 and 95% CI: 0.40–0.66, which is inferior to lncRNAs.\\nFurthermore, we developed logistic regression models for each lncRNA and commonly used tumor markers, and calculated the regression coefficients separately. We subsequently used the regression coefficients as weights to establish the combined index. Interestingly, both combined indices outperformed single lncRNA alone or traditional biomarkers alone in terms of ESCC detection (Table S4). Next, we plotted the ROC curves and compared AUC values by the DeLong's test.\\n23\\n Index I had a significantly higher AUC value than Index II (p = 0.0045) with an AUC value of 0.84 (95% CI: 0.74–0.91), showing remarkably better diagnostic performance (Figure 4e and Table S4; p = 0.0045). The above results suggested that the combination of these three lncRNAs exhibited superior discriminatory power than traditional biomarkers.\",\n \"We further analyzed the serum lncRNAs levels in 80 ESCC patients and matched healthy samples. Being consistent with the results of the training phase, the expression levels of lncRNAs were significantly higher in ESCC patients than in healthy individuals (Figure 5a). Then, we plotted ROC curves to compare the diagnostic performance between lncRNAs and classic biomarkers, with maximum AUC values of 0.86 and 0.54 for lncRNAs and classic biomarkers, respectively (Figure 5b,c). Table S4 showed in detail the AUC values and 95% CI of lncRNAs and tumor markers during the training phase.\\nValidation phase: A panel of serum lncRNAs‐based diagnostic indicators index I. (a) The expression levels of the three candidate lncRNAs in esophageal squamous cell carcinoma (ESCC) (n = 80) and control's serum samples (n = 80). (b) Receiver operating characteristic (ROC) curves of candidate lncRNAs to discriminate ESCC from healthy control. (c) ROC curves of traditional tumor markers. (d) Comparison of the diagnostic performance of Index I and Index II. (e) Kaplan–Meier survival curves of ESCC patients with low and high Index I.**p < 0.01.\\nBeing consistent with the results of the training phase, the lncRNAs‐based Index I obtained the greatest diagnostic performance with an AUC value of 0.86 (95% CI: 0.81–0.92) (Table S4). In addition, the AUC values of lncRNAs‐based Index I were significantly outperformed to that from Index II (Figure 5d and Table S4; p < 0.001).\\nTo investigate Index I related clinicopathological factors, we collected ESCC patients from the discovery, training, and validation phases and classified them into high and low levels using the median of the Index I index as the threshold value. As shown in Table 1, higher Index I levels were associated with greater depth of infiltration (p = 0.0267), lymph node metastasis (p = 0.0012), advanced tumor stage (p = 0.0040), and poor prognosis (Figure 5e; p < 0.001).\\nThe correlation between the serum Index I expression and clinicopathological factors of ESCC\\nAbbreviations: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\",\n \"We then analyzed the associations between 140 ESCC mortality and both three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) using a forward conditional logistic regression model. As shown in Table 2, in the univariable analysis, tumor size, RASSF8‐AS1 expression and LINC00958 expression were associated with ESCC mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression and CYFRA21‐1 expression were independently associated with ESCC mortality. The long‐term prognostic values of three serum‐derived lncRNAs (AC098818.2, RASSF8‐AS1 and LINC00958) and classic tumor markers (CEA125, CA199, CA153, SCC, NSE, CYFRA21‐1) were analyzed using a Cox regression model and the results are listed in Table 3. In the univariable analysis, tumor depth, lymph node metastasis and RASSF8‐AS1 expression were significantly associated with long‐term mortality. However, in the multivariable analysis, only RASSF8‐AS1 expression was independently associated with long‐term mortality. In conclusion, RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC.\\nAssociations between three serum‐derived lncRNAs, classic tumor markers and ESCC mortality\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; OR, odds ratio; —, no data available.\\nUnivariate and multivariate Cox analysis of overall survival in ESCC patients\\nAbbreviations: CI, confidence interval; ESCC, esophageal squamous cell carcinoma; HR, hazard ration; —, no data available.\",\n \"As lncRNA RASSF8‐AS1 showed good performance for ESCC diagnosis and RASSF8‐AS1 expression was an independent prognostic factor for both short‐ and long‐term outcomes of ESCC. We speculated that RASSF8‐AS1 plays an oncogenic role in ESCC. We then determined that higher RASSF8‐AS1 was associated with clinicopathological factors (Table 4), such as larger tumor size (p < 0.0001), greater depth of infiltration (p < 0.0001), lymph node metastasis (p = 0.0278), and advanced tumor stage (p = 0.0110). The expression of RASSF8‐AS1 tended to decrease significantly after tumor resection surgery in ESCC patients (Figure 4c), and RASSF8‐AS1 was found to be significantly higher in ESCC in the mid‐late stage than in stage I through the GEPIA database (Figure S3 D). The above data suggested that RASSF8‐AS1 plays a protumor role for ESCC.\\nThe correlation between the serum RASSF8‐AS1 expression and clinicopathological factors of ESCC\\nAbbreviation: ESCC, esophageal squamous cell carcinoma; —, no data available.\\nBold values means p‐value < 0.05.\\nSubsequently, we explored the oncogenic function of RASSF8‐AS1 by knocking down RASSF8‐AS1 in TE1 and TE13 cells (Figure 6a). In vitro assays showed that the knockdown of RASSF8‐AS1 inhibited cell proliferation (Figure 6b,c), motility and invasiveness (Figure 6d), respectively.\\nRASSF8‐AS1 promotes esophageal squamous cell carcinoma (ESCC) cell proliferation, migration and invasion in vitro. (a) The mRNA expression of RASSF8‐AS1 after transfection by two siRNAs. (b, c) Assessment of knockdown of RASSF8‐AS1 on cell growth of TE1 and TE13 cells. (d) Assessment of knockdown of RASSF8‐AS1 on cell migration and invasion. (e) The mRNA expression of RASSF8‐AS1 in normal tissues and ESCC tissues. (f) Kaplan–Meier survival curves of ESCC patients with low and high RASSF8‐AS1 expression. *p < 0.05, **p < 0.01, ***p < 0.001.\\nIn addition, we performed qRT‐PCR on 60 pairs of tumor and adjacent normal tissues, and found that RASSF8‐AS1 was likewise markedly elevated in ESCC tissues (Figure 6e). Next, we divided the 60 ESCC samples into RASSF8‐AS1 low expression group (n = 30) and high expression group (n = 30) according to the median level of RASSF8‐AS1 to explore the prognostic significance of RASSF8‐AS1 in ESCC. Kaplan–Meier analysis showed that patients with high RASSF8‐AS1 expression had a better dismal OS (Figure 6f).\\nFinally, we screened RASSF8‐AS1 targets by RNA‐sequencing and performed gene ontology (GO) and pathway enrichment analysis. The result found 303 positively regulated (Figure S4 C, D) and 430 negatively regulated targets (Figure 4e,f). GO results suggested RASSF8‐AS1 was positively involved, mainly in immune response and cell migration, and KEGG pathway analysis indicated the positively regulated targets mostly enriched in cell metabolism (Figure S4 E, F), all of which have been proven to be essential for tumor progression.\",\n \"Delayed diagnosis is the main cause for poor prognosis and rapid progression of ESCC. The five‐year survival rate for patients with early diagnosis of ESCC is >90%, while survival rate with late detection is <20%.\\n24\\n Endoscopic surveillance and biopsy examination are current applied approaches for ESCC diagnosis and monitoring, but are not popular because of the high cost and level of discomfort, while imaging technology is limited for the low sensitivity, thereby highlighting the urgent need to develop new, relatively noninvasive and reliable tools to improve the early diagnosis of ESCC.\\nThere are currently no widely accepted serum biomarkers for ESCC screening, surveillance for treatment response and recurrence. Traditional tumor biomarkers are inadequate in ESCC detection as they have poor sensitivity and specificity.\\nIn recent years, the emergence of novel liquid tumor markers has provided a new solution for the early screening of tumors. The discovery of circulating tumor cells (CTC),\\n25\\n circulating tumor cell DNA (ctDNA),\\n26\\n circulating cell‐free DNA (cfDNA),\\n27\\n and exosomes can further improve the efficacy of noninvasive screening. These humoral markers reflect the molecular and genetic information, as well as temporal and spatial heterogeneity of tumors, making the dynamic monitoring of tumor status possible. Among the currently acceptable noninvasive markers, the total amount of CTC is relatively low, while the specificity of ctDNA and cfDNA is insufficient, making the clinical application full of challenges. With the development of exosome isolation and enrichment technologies, exosome diagnosis has become a rising star in the field of liquid biopsy. Benefiting from the protection of the phospholipid bilayer on the surface of exosomes, the exosome contents have good stability, especially the RNAs can be protected by the membrane to avoid being degraded by RNase. The discovery of noncoding RNAs has enriched the understanding of tumorigenesis mechanisms. The role of exosome‐transported noncoding RNAs in tumor metastasis, drug resistance, and radio‐resistance across cells has received increasing attention. Studies have shown that the number of LncRNAs is huge, dozens of times that of currently known mRNAs, with high tissue specificity, which provides ideal conditions for the development of noninvasive screening. It has been reported that serum exosomal biomarkers may be derived from tumor tissues,\\n28\\n therefore in this study we also incorporated the upregulated lncRNAs from tissue microarray. The linear association between serum and tissue samples for the candidate lncRNAs confirmed the assumption. Finally, we identified a novel panel of exosomal lncRNAs for ESCC liquid biopsy through a multiphase study.\\nThe lncRNAs‐based composite Index I showed significantly higher diagnostic power than traditional tumor biomarker based composite Index II in two independent cohorts. Furthermore, the novel lncRNAs signature was stably detectable in serum, supporting the clinically applicable value. We also noted serum lncRNAs declined post‐surgery, indicating that Index I may reflect tumor burden of ESCC patients, which should be used as a follow‐up parameter for ESCC.\\nAmong the newly identified lncRNA panel, RASSF8‐AS1 showed the best diagnostic power for all ESCC patients, suggesting its oncogenic role in ESCC. Currently, the function of RASSF8‐AS1 has only been reported in laryngeal squamous cell carcinoma.\\n29\\n Our results showed its oncogenic role in ESCC for the first time and revealed the underlying mechanisms by target mRNA scanning. Similarly, the role of LIN00958 in ESCC has been recently reported, which also verified the oncogenic role of LIN00958, suggesting our results are reasonable.\\n30\\n\\n\\nWe also note the limitations of this study. First, we included only a relatively small number of patients and healthy controls, and the ESCC patients were diagnosed with cancer as they had already been diagnosed pathologically earlier, which should be verified in a much larger sample size. Second, we conducted a retrospective, single‐center, cross‐sectional study, which may have introduced unavoidable selection bias. A prospective multicenter study is needed to further confirm the diagnostic efficiency of Index I, and to check the superiority to traditional biomarkers. Next, how lncRNA enters the exosome, how lncRNA encapsulated exosomes play a role in tumor progress and distant migration call for more attention. In recent years, the role of exosomes in mediating distant tumor metastasis has gained emphasis,\\n31\\n, \\n32\\n and our study revealed high expression of serum lncRNA RASSF8‐AS1 was also correlated with metastasis, and our sequencing results suggested its involvement in immune response and cell metabolism. How our newly discovered exosomal lncRNAs play a role in distant metastasis of ESCC deserves further investigation.\",\n \"This research was supported by the National Natural Science Foundation for Youth of China (no. 81702444), China Postdoctoral Science Foundation 2018M643884 and 2020T130128ZX.\",\n \"All authors declare no conflict of interest.\",\n \"\\nTable S1 The sequences of qRT‐PCR primers.\\n\\nTable S2 The sequences of siRNAs.\\n\\nTable S3 Demographic and clinicopathologic characteristics of the study subjects.\\n\\nTable S4 Results of ROC curves for individual biomarkers.\\nClick here for additional data file.\\n\\nFigure S1 Characterization of exosomes isolated from serum samples. (A) TEM images of exosomes. HV = 100.0 kV, Direct Mag: 100000× (scale bars 100 nm). Exosomes were winkled oval or spherical in shape under TEM. (B) Size and concentration distribution of exosomes were analyzed by NTA.\\nClick here for additional data file.\\n\\nFigure S2 Bioinformatic analysis of differential mRNAs. (A) Hierarchical clustering of the high‐throughput microarray performed for serum mRNA. (B–D) Gene ontology enrichments of biological process (BP), Cellular component (CC) and molecular function (MF) for differential mRNAs.\\nClick here for additional data file.\\n\\nFigure S3 Big data mining for the candidate lncRNAs. (A–C) Expression of serum exosomal lncRNAs in pan‐cancer compared with healthy donors. (D) Tissues with higher stages exhibit roughly gradually elevated expression of RASSF8‐AS1. *p < 0.05, **p < 0.01, ***p < 0.001.\\nClick here for additional data file.\\n\\nFigure S4 RNA‐sequencing results of RSSF8‐AS1 targets. (A) Volcano plots for the differential mRNAs. (B) Hierarchical Clustering for the differentially expressed RNA. (C, D) Gene Oncology and pathway enrichment analysis for the positively regulated mRNAs of RASSF8‐AS1. (E, F) Oncology and pathway enrichment analysis for the negatively regulated mRNAs of RASSF8‐AS1.\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,null,null,null,null,null,null,null,null,null,"results",null,null,null,null,null,null,null,"discussion",null,"COI-statement","supplementary-material"],"string":"[\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n \"COI-statement\",\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["circulating lncRNAs","esophageal squamous cell carcinoma","exosome; liquid biopsy","prognosis"],"string":"[\n \"circulating lncRNAs\",\n \"esophageal squamous cell carcinoma\",\n \"exosome; liquid biopsy\",\n \"prognosis\"\n]"}}},{"rowIdx":370,"cells":{"title":{"kind":"string","value":"Gothenburg Empowerment Scale (GES): psychometric properties and measurement invariance in adults with congenital heart disease from Belgium, Norway and South Korea."},"pmid":{"kind":"string","value":"36266608"},"background_abstract":{"kind":"string","value":"Patient empowerment is associated with improvements in different patient-reported and clinical outcomes. However, despite being widely researched, high quality and theoretically substantiated disease-generic measures of patient empowerment are lacking. The few good instruments that are available have not reported important psychometric properties, including measurement invariance. The aim of this study was to assess the psychometric properties of the 15-item Gothenburg Empowerment Scale (GES), with a particular focus on measurement invariance of the GES across individuals from three countries."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Adults with congenital heart disease from Belgium, Norway and South Korea completed the GES and other patient-reported outcomes as part of an international, cross-sectional, descriptive study called APPROACH-IS II. The scale's content (missing data) and factorial validity (confirmatory factor analyses), measurement invariance (multi-group confirmatory factor analyses), responsiveness (floor and ceiling effects) and reliability (internal consistency) were assessed."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Content validity, responsiveness and reliability were confirmed. Nonetheless, metric but not scalar measurement invariance was supported when including the three countries, possibly because the scale performed differently in the sample from South Korea. A second set of analyses supported partial scalar invariance for a sample that was limited to Norway and Belgium."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Our study offers preliminary evidence that GES is a valid and reliable measure of patient empowerment in adults with congenital heart disease. However, cross-country comparisons must be made with caution, given the scale did not perform equivalently across the three countries."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Adult","Humans","Psychometrics","Reproducibility of Results","Surveys and Questionnaires","Cross-Sectional Studies","Belgium","Quality of Life","Factor Analysis, Statistical","Republic of Korea","Heart Defects, Congenital"],"string":"[\n \"Adult\",\n \"Humans\",\n \"Psychometrics\",\n \"Reproducibility of Results\",\n \"Surveys and Questionnaires\",\n \"Cross-Sectional Studies\",\n \"Belgium\",\n \"Quality of Life\",\n \"Factor Analysis, Statistical\",\n \"Republic of Korea\",\n \"Heart Defects, Congenital\"\n]"},"pmcid":{"kind":"string","value":"9583060"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Content validity [SUBSECTION] The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\nThe proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\n[SUBTITLE] Factorial validity [SUBSECTION] The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES\nThe five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES\n[SUBTITLE] Measurement invariance [SUBSECTION] The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nThe configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n[SUBTITLE] Internal consistency [SUBSECTION] The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\nThe Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\n[SUBTITLE] Responsiveness [SUBSECTION] The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores\nThe mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"This study provides evidence on the GES’s validity, reliability, and responsiveness in adults with CHD. GES seems to be valid and reliable for the sample of Belgium and Norway. However, this is not the case for South Korea, because shared decision-making seems to have a different meaning in Asian cultures. Hence, cross-cultural comparisons using GES should be made cautiously. It is possible that in countries who are more culturally different from Norway or Belgium, the scale might not perform as well."},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Development of GES","Design","Sample","Procedure","Statistical analyses","Ethics and informed consent","Content validity","Factorial validity","Measurement invariance","Measurement invariance: norwegian and belgian samples","Internal consistency","Responsiveness","Methodological considerations"],"string":"[\n \"Background\",\n \"Methods\",\n \"Development of GES\",\n \"Design\",\n \"Sample\",\n \"Procedure\",\n \"Statistical analyses\",\n \"Ethics and informed consent\",\n \"Content validity\",\n \"Factorial validity\",\n \"Measurement invariance\",\n \"Measurement invariance: norwegian and belgian samples\",\n \"Internal consistency\",\n \"Responsiveness\",\n \"Methodological considerations\"\n]"},"other_sections_texts":{"kind":"list like","value":["Patient empowerment is a concept associated with increasing peoples’ ability to manage their condition and their lives [1]. It initially emerged in the health promotion field and later demonstrated relevance in the care of persons with chronic conditions (CCs) [2]. It has been suggested that by increasing patient empowerment, patients may eventually become more actively engaged in their care, develop more disease-related knowledge, and improve their quality of life and well-being [3].\nThere is a considerable amount of research on patient empowerment, both theoretically and empirically [4]. However, to date, there is no consensus on how to define patient empowerment or which instrument is best at capturing this construct. There are approximately 30 different definitions available and more than 40 instruments [4]. Moreover, a detailed assessment of current instruments has indicated that a vast majority lack a clear theoretical ground and have poor validity and reliability [5, 6].\nGiven the methodological limitations of existing instruments, a new measure, the Gothenburg Young Persons Empowerment Scale (GYPES), was developed [7]. Scale construction reflected the theoretical work of Small and colleagues who defined patient empowerment as “an enabling process or outcome arising from communication with the healthcare professional and a mutual sharing of resources over information relating to illness, which enhances the patients’ feelings of control, self-efficacy, coping abilities and ability to achieve change over their condition” [8]. From semi-structured interviews with adults with CCs, they concluded that patient empowerment comprises five dimensions [8]. First, knowledge and understanding, related to the level of knowledge patients need to manage their illness and lives. Second, personal control, given that each patient should have the ability to manage their disease. Third, identity, which entails how much the illness influences the patients’ lives and their sense of self. Fourth, shared decision-making, the ability and possibility to make decisions together with the healthcare provider. Fifth, enabling others, referring to the ability to share experiences and coping strategies with other persons who are experiencing a similar situation.\nGYPES was developed for use with adolescents with CCs because at the time there were no existing instruments developed specifically for this age group. The scale’s content and factorial validity, internal consistency, and responsiveness were evaluated in two samples of young persons with congenital heart disease (CHD) and type I diabetes. GYPES was found to have good psychometric properties in these groups [7].\nGiven that the GYPES was found to be valid and reliable for its use with adolescents with CCs, and because it was based on sound conceptual grounds, we developed a slightly modified version to measure empowerment in adults with CCs: the Gothenburg Empowerment Scale (GES). In the evaluation of the scale’s psychometric properties reported herein, we also chose to investigate measurement invariance, given potential use of the GES in international studies. Measurement invariance, also known as measurement equivalence, is often neglected in psychometric evaluation [9], yet it is important to investigate whether an instrument measures the same construct in different contexts. If invariance is not supported, this suggests that the instrument triggers different response mechanisms in different groups, making score comparisons invalid [10]. Invariance is particularly important if cross-cultural comparisons are to be made [11].\nTherefore, the aim of this study was to assess the psychometric properties of the GES, including the level of measurement invariance among adults with CHD from different cultures.","[SUBTITLE] Development of GES [SUBSECTION] Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\nDevelopment of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\n[SUBTITLE] Design [SUBSECTION] This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\nThis methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\n[SUBTITLE] Sample [SUBSECTION] APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)\nAPPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)\n[SUBTITLE] Procedure [SUBSECTION] Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\nParticipants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\n[SUBTITLE] Statistical analyses [SUBSECTION] The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\nThe psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\n[SUBTITLE] Ethics and informed consent [SUBSECTION] APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\nAPPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.","Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.","This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.","APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)","Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.","The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.","APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.","The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.","The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES","The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030","As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030","The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.","The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores","There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited."],"string":"[\n \"Patient empowerment is a concept associated with increasing peoples’ ability to manage their condition and their lives [1]. It initially emerged in the health promotion field and later demonstrated relevance in the care of persons with chronic conditions (CCs) [2]. It has been suggested that by increasing patient empowerment, patients may eventually become more actively engaged in their care, develop more disease-related knowledge, and improve their quality of life and well-being [3].\\nThere is a considerable amount of research on patient empowerment, both theoretically and empirically [4]. However, to date, there is no consensus on how to define patient empowerment or which instrument is best at capturing this construct. There are approximately 30 different definitions available and more than 40 instruments [4]. Moreover, a detailed assessment of current instruments has indicated that a vast majority lack a clear theoretical ground and have poor validity and reliability [5, 6].\\nGiven the methodological limitations of existing instruments, a new measure, the Gothenburg Young Persons Empowerment Scale (GYPES), was developed [7]. Scale construction reflected the theoretical work of Small and colleagues who defined patient empowerment as “an enabling process or outcome arising from communication with the healthcare professional and a mutual sharing of resources over information relating to illness, which enhances the patients’ feelings of control, self-efficacy, coping abilities and ability to achieve change over their condition” [8]. From semi-structured interviews with adults with CCs, they concluded that patient empowerment comprises five dimensions [8]. First, knowledge and understanding, related to the level of knowledge patients need to manage their illness and lives. Second, personal control, given that each patient should have the ability to manage their disease. Third, identity, which entails how much the illness influences the patients’ lives and their sense of self. Fourth, shared decision-making, the ability and possibility to make decisions together with the healthcare provider. Fifth, enabling others, referring to the ability to share experiences and coping strategies with other persons who are experiencing a similar situation.\\nGYPES was developed for use with adolescents with CCs because at the time there were no existing instruments developed specifically for this age group. The scale’s content and factorial validity, internal consistency, and responsiveness were evaluated in two samples of young persons with congenital heart disease (CHD) and type I diabetes. GYPES was found to have good psychometric properties in these groups [7].\\nGiven that the GYPES was found to be valid and reliable for its use with adolescents with CCs, and because it was based on sound conceptual grounds, we developed a slightly modified version to measure empowerment in adults with CCs: the Gothenburg Empowerment Scale (GES). In the evaluation of the scale’s psychometric properties reported herein, we also chose to investigate measurement invariance, given potential use of the GES in international studies. Measurement invariance, also known as measurement equivalence, is often neglected in psychometric evaluation [9], yet it is important to investigate whether an instrument measures the same construct in different contexts. If invariance is not supported, this suggests that the instrument triggers different response mechanisms in different groups, making score comparisons invalid [10]. Invariance is particularly important if cross-cultural comparisons are to be made [11].\\nTherefore, the aim of this study was to assess the psychometric properties of the GES, including the level of measurement invariance among adults with CHD from different cultures.\",\n \"[SUBTITLE] Development of GES [SUBSECTION] Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\\nDevelopment of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\\n[SUBTITLE] Design [SUBSECTION] This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\\nThis methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\\n[SUBTITLE] Sample [SUBSECTION] APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\\nAPPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\\n[SUBTITLE] Procedure [SUBSECTION] Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\\nParticipants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\\n[SUBTITLE] Statistical analyses [SUBSECTION] The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\\nThe psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\\n[SUBTITLE] Ethics and informed consent [SUBSECTION] APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\\nAPPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\",\n \"Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\",\n \"This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\",\n \"APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\",\n \"Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\",\n \"The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\",\n \"APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\",\n \"The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\",\n \"The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\\n\\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\\nFirst-order factor\\n\\nSecond-order factor\\n\\nKnowledge and Understanding\\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\\nPersonal Control\\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\\nIdentity\\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\\nShared decision-making\\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\\nEnabling Others\\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\\n\\nProportion of missing values and factor loadings for GES\",\n \"The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\",\n \"As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\",\n \"The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\",\n \"The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\\n\\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\\nCronbach’s alpha values\\n\\nKnowledge and understanding\\n0.7300.7210.7590.728\\nPersonal control\\n0.6930.6330.6360.751\\nIdentity\\n0.6800.6820.7510.636\\nShared decision-making\\n0.7740.8330.8040.491\\nEnabling others\\n0.8630.8720.9050.782\\nOverall scale\\n0.8730.8780.8710.854\\nComposite reliability\\n\\nKnowledge and understanding\\n0.7530.7520.7790.732\\nPersonal control\\n0.7020.6500.6770.753\\nIdentity\\n0.6840.6820.7700.657\\nShared decision-making\\n0.5590.8350.8050.470\\nEnabling others\\n0.8640.8760.9100.782\\nOverall scale\\n0.9290.9420.9510.917\\nMean patient empowerment scores\\n\\nKnowledge and understanding\\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\\nPersonal control\\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\\nIdentity\\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\\nShared decision-making\\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\\nEnabling others\\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\\nOverall scale\\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\\n\\nCronbach’s alpha values, composite reliability and mean empowerment scores\",\n \"There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Development of GES","Design","Sample","Procedure","Statistical analyses","Ethics and informed consent","Results","Content validity","Factorial validity","Measurement invariance","Measurement invariance: norwegian and belgian samples","Internal consistency","Responsiveness","Discussion","Methodological considerations","Conclusion"],"string":"[\n \"Background\",\n \"Methods\",\n \"Development of GES\",\n \"Design\",\n \"Sample\",\n \"Procedure\",\n \"Statistical analyses\",\n \"Ethics and informed consent\",\n \"Results\",\n \"Content validity\",\n \"Factorial validity\",\n \"Measurement invariance\",\n \"Measurement invariance: norwegian and belgian samples\",\n \"Internal consistency\",\n \"Responsiveness\",\n \"Discussion\",\n \"Methodological considerations\",\n \"Conclusion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Patient empowerment is a concept associated with increasing peoples’ ability to manage their condition and their lives [1]. It initially emerged in the health promotion field and later demonstrated relevance in the care of persons with chronic conditions (CCs) [2]. It has been suggested that by increasing patient empowerment, patients may eventually become more actively engaged in their care, develop more disease-related knowledge, and improve their quality of life and well-being [3].\nThere is a considerable amount of research on patient empowerment, both theoretically and empirically [4]. However, to date, there is no consensus on how to define patient empowerment or which instrument is best at capturing this construct. There are approximately 30 different definitions available and more than 40 instruments [4]. Moreover, a detailed assessment of current instruments has indicated that a vast majority lack a clear theoretical ground and have poor validity and reliability [5, 6].\nGiven the methodological limitations of existing instruments, a new measure, the Gothenburg Young Persons Empowerment Scale (GYPES), was developed [7]. Scale construction reflected the theoretical work of Small and colleagues who defined patient empowerment as “an enabling process or outcome arising from communication with the healthcare professional and a mutual sharing of resources over information relating to illness, which enhances the patients’ feelings of control, self-efficacy, coping abilities and ability to achieve change over their condition” [8]. From semi-structured interviews with adults with CCs, they concluded that patient empowerment comprises five dimensions [8]. First, knowledge and understanding, related to the level of knowledge patients need to manage their illness and lives. Second, personal control, given that each patient should have the ability to manage their disease. Third, identity, which entails how much the illness influences the patients’ lives and their sense of self. Fourth, shared decision-making, the ability and possibility to make decisions together with the healthcare provider. Fifth, enabling others, referring to the ability to share experiences and coping strategies with other persons who are experiencing a similar situation.\nGYPES was developed for use with adolescents with CCs because at the time there were no existing instruments developed specifically for this age group. The scale’s content and factorial validity, internal consistency, and responsiveness were evaluated in two samples of young persons with congenital heart disease (CHD) and type I diabetes. GYPES was found to have good psychometric properties in these groups [7].\nGiven that the GYPES was found to be valid and reliable for its use with adolescents with CCs, and because it was based on sound conceptual grounds, we developed a slightly modified version to measure empowerment in adults with CCs: the Gothenburg Empowerment Scale (GES). In the evaluation of the scale’s psychometric properties reported herein, we also chose to investigate measurement invariance, given potential use of the GES in international studies. Measurement invariance, also known as measurement equivalence, is often neglected in psychometric evaluation [9], yet it is important to investigate whether an instrument measures the same construct in different contexts. If invariance is not supported, this suggests that the instrument triggers different response mechanisms in different groups, making score comparisons invalid [10]. Invariance is particularly important if cross-cultural comparisons are to be made [11].\nTherefore, the aim of this study was to assess the psychometric properties of the GES, including the level of measurement invariance among adults with CHD from different cultures.","[SUBTITLE] Development of GES [SUBSECTION] Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\nDevelopment of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\n[SUBTITLE] Design [SUBSECTION] This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\nThis methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\n[SUBTITLE] Sample [SUBSECTION] APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)\nAPPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)\n[SUBTITLE] Procedure [SUBSECTION] Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\nParticipants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\n[SUBTITLE] Statistical analyses [SUBSECTION] The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\nThe psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\n[SUBTITLE] Ethics and informed consent [SUBSECTION] APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\nAPPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.","Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.","This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.","APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\n\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\nAge at inclusion (median, (Q1;Q3))\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\nSex (n (%))\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\nEducation (n (%))\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\nCHD complexity (n (%))\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\n\nSociodemographic and clinical aspects\n\nSex (n (%))\n\nFemale\nMale\n430 (51.0)\n413 (49.0)\n240 (48.3)\n253 (50.9)\n86 (60.1)\n57 (39.9)\n104(50.2)\n103 (49.8)\n\nEducation (n (%))\n\nNo high school education\nHigh school\nBachelor’s degree\nMaster’s degree or higher\n447 (5.2)\n346 (41.2)\n258 (30.8)\n190 (22.6)\n29 (5.9)\n194 (39.6)\n160 (32.7)\n106 (21.6)\n10 (6.9)\n60 (41.7)\n33 (22.9)\n41 (28.5)\n5 (2.4)\n92 (44.9)\n65 (31.7)\n43 (21.0)\n\nCHD complexity (n (%))\n\nMild\nModerate\nSevere\n116 (13.8)\n473 (56.4)\n249 (29.7)\n98 (19.7)\n313 (63.0)\n86 (17.3)\n8 (6.0)\n82 (61.7)\n43 (32.3)\n10 (4.8)\n78 (37.5)\n120 (57.7)","Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.","The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.","APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.","[SUBTITLE] Content validity [SUBSECTION] The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\nThe proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\n[SUBTITLE] Factorial validity [SUBSECTION] The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES\nThe five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES\n[SUBTITLE] Measurement invariance [SUBSECTION] The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nThe configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n[SUBTITLE] Internal consistency [SUBSECTION] The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\nThe Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\n[SUBTITLE] Responsiveness [SUBSECTION] The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores\nThe mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores","The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.","The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\n\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\nFirst-order factor\n\nSecond-order factor\n\nKnowledge and Understanding\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\nPersonal Control\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\nIdentity\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\nShared decision-making\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\nEnabling Others\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\n\nProportion of missing values and factor loadings for GES","The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030","As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\n\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\nComplete sample (Belgium, Norway and South Korea)\n\nConfigural invariance\na\n772.074 (255)0.8990.0850.063\nConfigural invariance\nb\n695.911 (252)0.9130.0790.061\nMetric invariance\nb\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\nScalar invariance\nb\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\nPartial scalar invariance with intercept of item 9 free\nb\n886.943 (296)0.8850.0840.072\nPartial scalar invariance with intercept of items 9 and 4 free\nb\n856.568 (294)0.8900.0820.070\nPartial scalar invariance with intercept of items 9, 4 and 12 free\nb\n827.343 (292)0.8960.0800.069\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\nb\n813.205 (290)0.8980.0800.069\nPartial sample (Belgium and Norway)\n\nConfigural invariance\na\n463.510 (170)0.9280.0740.057\nMetric invariance\na\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\nScalar invariance\na\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\nPartial scalar invariance with one intercept free\nc\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\n\nModel fit indexes of multi-group confirmatory factor analyses\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030","The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.","The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\n\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\nCronbach’s alpha values\n\nKnowledge and understanding\n0.7300.7210.7590.728\nPersonal control\n0.6930.6330.6360.751\nIdentity\n0.6800.6820.7510.636\nShared decision-making\n0.7740.8330.8040.491\nEnabling others\n0.8630.8720.9050.782\nOverall scale\n0.8730.8780.8710.854\nComposite reliability\n\nKnowledge and understanding\n0.7530.7520.7790.732\nPersonal control\n0.7020.6500.6770.753\nIdentity\n0.6840.6820.7700.657\nShared decision-making\n0.5590.8350.8050.470\nEnabling others\n0.8640.8760.9100.782\nOverall scale\n0.9290.9420.9510.917\nMean patient empowerment scores\n\nKnowledge and understanding\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\nPersonal control\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\nIdentity\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\nShared decision-making\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\nEnabling others\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\nOverall scale\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\n\nCronbach’s alpha values, composite reliability and mean empowerment scores","Recent research highlights the need to improve and evaluate patient empowerment in persons with CCs [4]. However, the availability of instruments with a strong theoretical background and/or acceptable psychometric properties is limited [6]. Therefore, the present study evaluated a slightly modified version of a previously validated patient empowerment in a group of adults with CHD. Results indicate that GES proved reliable, meaning the scale’s items altogether are consistent. Additionally, no floor or ceiling effects were identified, which is indirectly related to the scale’s sensitivity to measure change [22]. Whereas these results are in favor of the GES, a psychometric assessment of the scale’s validity revealed that the scale appears valid in two of the three included countries.\nResults from the CFA of the entire sample indicate the five-factor structure fits the data well. However, due to the lack of measurement invariance, CFAs were performed separately for patients from each country. Analyses revealed that an acceptable model fit was only achieved for the samples from Belgium and Norway. For the South Korean data, the CFI and the RMSEA were above the acceptable ranges although factor loadings from the “shared decision-making” dimension were low. This is an indication that the relations among the latent constructs of GES are not in line with the data, i.e., the data do not support the model [24]. Even though it is common in CFA to conduct an evaluation of modification indexes to improve model fit, changes should be theoretically justified [25]. In this case, we could identify no strong theoretical reasons to allow for model changes that could potentially improve the model fit.\nThe poorly fitted model to the South Korean data might be the reason why invariance was not achieved when comparing the three countries. A lack of invariance suggests that patient empowerment may have a different meaning in different countries. Therefore, making mean comparisons of this construct across countries using the GES should be made with caution. Nonetheless, while cross-country comparisons cannot be made, it is possible to compare how the GES relates to other constructs (i.e., variables) within countries. For example, one could undertake a study to compare how the GES relates to patient functioning in Belgium, vs. how the GES relates to patient functioning in South Korea. Another potential research aim would be to determine whether changes in GES relate to psychological distress and whether these changes have the same magnitude in Belgium, Norway and South Korea.\nThe absence of invariance across the three countries can be an indicator of patient empowerment being interpreted differently across countries. Perhaps individuals in South Korea have different interpretations and values of patient empowerment compared to patients in Belgium and Norway, two European countries that are likely more culturally similar. It has been suggested that patient empowerment is determined by its context and that individuals within a particular environment, organization or country may have a different perception of empowerment [26]. Hence, a measure that fits all persons (or contexts), might be hard to develop. Findings from the present study partially support this notion. It is plausible to consider that different countries (and cultures) have a different understanding of patient empowerment domains, values, and skills.\nEvidence on how patient empowerment may differ between countries is limited. However, authors of a systematic review concluded that although patients do share many perspectives of this construct, there might be variations on how certain aspects of patient empowerment are understood and research has yet to address structural elements of patient empowerment [27].\nConsideration of structural elements is relevant for the current study because the “shared decision-making” dimension had the lowest factor loadings for the South Korean sample, as well as poor reliability. Shared decision-making involves healthcare professionals and patients working together on a care plan [28]. It entails patients who are willing to participate as well as clinicians who want to collaborate with the patient. Hence, it is greatly influenced by the healthcare structure. Within Asian cultures, for example, it is more commonly believed that patients prefer to take a less independent role and that a family-centered (rather than individual-centered) approach towards decision-making is preferred [29–31]. It has been suggested that shared decision-making reflects values associated with western cultures and that this might be different in other cultures [28, 31, 32]. The value placed on shared decision-making and the structure of the healthcare system might therefore potentially impact the value of patient empowerment as perceived by study participants. Therefore, the comparison of this construct between different cultures and regions of the world should be made with caution.\nAlthough the GES appears to be a valid and reliable instrument for the Norwegian and Belgian sample, it is worth noting that future research should evaluate whether the low psychometric properties found in the South Korean sample are indeed applicable to other Asian or non-European countries. Comparatives studies evaluating this construct (and similar constructs) between Western and Asian countries are needed to comprehend this phenomenon, the types of research questions that can be addressed and the conclusions that can be drawn when evaluating outcomes such as patient empowerment. Such studies could be undertaken once the data collection from APPROACH-IS II is finalized.\n[SUBTITLE] Methodological considerations [SUBSECTION] There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.\nThere are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.","There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.","This study provides evidence on the GES’s validity, reliability, and responsiveness in adults with CHD. GES seems to be valid and reliable for the sample of Belgium and Norway. However, this is not the case for South Korea, because shared decision-making seems to have a different meaning in Asian cultures. Hence, cross-cultural comparisons using GES should be made cautiously. It is possible that in countries who are more culturally different from Norway or Belgium, the scale might not perform as well."],"string":"[\n \"Patient empowerment is a concept associated with increasing peoples’ ability to manage their condition and their lives [1]. It initially emerged in the health promotion field and later demonstrated relevance in the care of persons with chronic conditions (CCs) [2]. It has been suggested that by increasing patient empowerment, patients may eventually become more actively engaged in their care, develop more disease-related knowledge, and improve their quality of life and well-being [3].\\nThere is a considerable amount of research on patient empowerment, both theoretically and empirically [4]. However, to date, there is no consensus on how to define patient empowerment or which instrument is best at capturing this construct. There are approximately 30 different definitions available and more than 40 instruments [4]. Moreover, a detailed assessment of current instruments has indicated that a vast majority lack a clear theoretical ground and have poor validity and reliability [5, 6].\\nGiven the methodological limitations of existing instruments, a new measure, the Gothenburg Young Persons Empowerment Scale (GYPES), was developed [7]. Scale construction reflected the theoretical work of Small and colleagues who defined patient empowerment as “an enabling process or outcome arising from communication with the healthcare professional and a mutual sharing of resources over information relating to illness, which enhances the patients’ feelings of control, self-efficacy, coping abilities and ability to achieve change over their condition” [8]. From semi-structured interviews with adults with CCs, they concluded that patient empowerment comprises five dimensions [8]. First, knowledge and understanding, related to the level of knowledge patients need to manage their illness and lives. Second, personal control, given that each patient should have the ability to manage their disease. Third, identity, which entails how much the illness influences the patients’ lives and their sense of self. Fourth, shared decision-making, the ability and possibility to make decisions together with the healthcare provider. Fifth, enabling others, referring to the ability to share experiences and coping strategies with other persons who are experiencing a similar situation.\\nGYPES was developed for use with adolescents with CCs because at the time there were no existing instruments developed specifically for this age group. The scale’s content and factorial validity, internal consistency, and responsiveness were evaluated in two samples of young persons with congenital heart disease (CHD) and type I diabetes. GYPES was found to have good psychometric properties in these groups [7].\\nGiven that the GYPES was found to be valid and reliable for its use with adolescents with CCs, and because it was based on sound conceptual grounds, we developed a slightly modified version to measure empowerment in adults with CCs: the Gothenburg Empowerment Scale (GES). In the evaluation of the scale’s psychometric properties reported herein, we also chose to investigate measurement invariance, given potential use of the GES in international studies. Measurement invariance, also known as measurement equivalence, is often neglected in psychometric evaluation [9], yet it is important to investigate whether an instrument measures the same construct in different contexts. If invariance is not supported, this suggests that the instrument triggers different response mechanisms in different groups, making score comparisons invalid [10]. Invariance is particularly important if cross-cultural comparisons are to be made [11].\\nTherefore, the aim of this study was to assess the psychometric properties of the GES, including the level of measurement invariance among adults with CHD from different cultures.\",\n \"[SUBTITLE] Development of GES [SUBSECTION] Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\\nDevelopment of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\\n[SUBTITLE] Design [SUBSECTION] This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\\nThis methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\\n[SUBTITLE] Sample [SUBSECTION] APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\\nAPPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\\n[SUBTITLE] Procedure [SUBSECTION] Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\\nParticipants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\\n[SUBTITLE] Statistical analyses [SUBSECTION] The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\\nThe psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\\n[SUBTITLE] Ethics and informed consent [SUBSECTION] APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\\nAPPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\",\n \"Development of the GES was modeled after the 15-item GYPES, which includes three items each to assess the five dimensions of patient empowerment: (1) personal control; (2) knowledge and understanding; (3) identity; (4) shared decision-making; and (5) enabling others [8]. Scale items are purposefully disease-generic, to facilitate use within different populations of CCs. Originally created in English, the GYPES has currently been translated to other languages including Swedish, Dutch, Mandarin, and Turkish. Additional information on the development and evaluation of GYPES has been published [7].\\nThe GES includes the same five dimensions and number of items as GYPES. The only modification was changing the term “young persons” by “persons” throughout the scale. To translate the GES to Dutch, Norwegian and Korean for the present study, the translation process followed the guidelines from the World Health Organization [12]. This process entailed a forward-backward translation, pre-testing the translated questionnaires in a few patients, proofreading the questionnaire and finalizing the translation. To assure consistency, substantial changes from the English version were not permitted.\",\n \"This methodological study is part of a larger international, cross-sectional descriptive study known as APPROACH-IS II, which aims to increase the understanding of patient-reported outcomes in adults with CHD, by enrolling adults with CHD from 32 different countries across the world [13]. APPROACH-IS II is registered at ClinicalTrials.gov: NCT04902768.\",\n \"APPROACH-IS II is enrolling participants from over 50 adult CHD centers worldwide. Data collection is scheduled to be finished by the end of August 2022. For the present study, data from participants of Belgium, Norway and South Korea were included, since the data collection of these three countries was undertaken before the covid-19 pandemic.\\nParticipants were eligible for the study if they fulfilled the following criteria: (i) diagnosis of CHD; (ii) aged 18 years or older at the moment of inclusion; (iii) diagnosed before the age of 10 years, (because we wanted participants to have experience living with CHD); (iv) in follow-up at an adult CHD center or included in a national/regional registry; and (v) having the physical, cognitive and language abilities to complete the self-report questionnaires. Patients with prior heart transplantation were ineligible.\\nIn the present study, there are a total of 850 people enrolled, 497 were from Belgium, 144 from Norway, and 209 from South Korea. The median age for the total sample was 30 years, and the proportion of men and women was fairly equal. Moderate CHD was found in 56.4% of the participants. Demographic and clinical information of the included participants is detailed in Table 1. There were significant differences in age (p < 0.001, eta square: 0.044) and complexity (p < 0.001, Cramer’s V: 0.227) between the samples, with moderate effect sizes. Indeed, the Norwegian sample was slightly older, and the proportion of severe CHD was larger in South Korea.\\n\\nTable 1Sociodemographic and clinical aspectsTotal Sample(n = 850)Belgium(n = 497)Norway(n = 144)South Korea(n = 209)\\nAge at inclusion (median, (Q1;Q3))\\n30.0 (27;43)29.0 (27;38)40.0 (30;53)31.0 (24;45)\\nSex (n (%))\\nFemaleMale430 (51.0)413 (49.0)240 (48.3)253 (50.9)86 (60.1)57 (39.9)104(50.2)103 (49.8)\\nEducation (n (%))\\nNo high school educationHigh schoolBachelor’s degreeMaster’s degree or higher447 (5.2)346 (41.2)258 (30.8)190 (22.6)29 (5.9)194 (39.6)160 (32.7)106 (21.6)10 (6.9)60 (41.7)33 (22.9)41 (28.5)5 (2.4)92 (44.9)65 (31.7)43 (21.0)\\nCHD complexity (n (%))\\nMildModerateSevere116 (13.8)473 (56.4)249 (29.7)98 (19.7)313 (63.0)86 (17.3)8 (6.0)82 (61.7)43 (32.3)10 (4.8)78 (37.5)120 (57.7)\\n\\nSociodemographic and clinical aspects\\n\\nSex (n (%))\\n\\nFemale\\nMale\\n430 (51.0)\\n413 (49.0)\\n240 (48.3)\\n253 (50.9)\\n86 (60.1)\\n57 (39.9)\\n104(50.2)\\n103 (49.8)\\n\\nEducation (n (%))\\n\\nNo high school education\\nHigh school\\nBachelor’s degree\\nMaster’s degree or higher\\n447 (5.2)\\n346 (41.2)\\n258 (30.8)\\n190 (22.6)\\n29 (5.9)\\n194 (39.6)\\n160 (32.7)\\n106 (21.6)\\n10 (6.9)\\n60 (41.7)\\n33 (22.9)\\n41 (28.5)\\n5 (2.4)\\n92 (44.9)\\n65 (31.7)\\n43 (21.0)\\n\\nCHD complexity (n (%))\\n\\nMild\\nModerate\\nSevere\\n116 (13.8)\\n473 (56.4)\\n249 (29.7)\\n98 (19.7)\\n313 (63.0)\\n86 (17.3)\\n8 (6.0)\\n82 (61.7)\\n43 (32.3)\\n10 (4.8)\\n78 (37.5)\\n120 (57.7)\",\n \"Participants completed a set of self-reported questionnaires, including the GES, that were administered during outpatient clinic visits (in Norway, South Korea and Belgium) and/or mailed to their home (in Belgium). Data for this sub-study were collected between August 2019 and February 2020, hence before the COVID-19 pandemic emerged.\",\n \"The psychometric evaluation of GES entailed an assessment of the content validity, factorial validity, measurement invariance, internal consistency and responsiveness. Content validity was assessed by the proportion of missing values and invalid scores [7]. While this is not a common approach for evaluating content validity, missing values can be considered an indicator of whether the items in the scale are perceived as relevant by the participants and whether the items are intelligible. Less than 5% missing data was deemed acceptable in this study.\\nFactorial validity was evaluated through confirmatory factor analyses (CFA) to assess the hypothesized five-factor structure of the scale and the overall factor of patient empowerment [14]. CFA was performed for the entire sample. A good model fit was obtained if the comparative fit index (CFI) was > 0.90, standardized root mean square residual (SRMR) < 0.08 and root mean square error of approximation (RMSEA) < 0.08 [15]. These fit indices were chosen based on Kline’s suggestion for fit evaluation of CFA [14]. Standardized factor loadings are also reported.\\nMeasurement invariance was examined through multigroup CFA (MGCFA) across countries [11]. The MGCFA included three models: (1) configural model (i.e., separates the sample into three subgroups, but no parameter constraints are imposed); (2) metric model (i.e., constrains the factor loadings to be equal across subgroups); and (3) scalar model (i.e., constrains the factor loadings and the item intercepts to be equal across subgroups) [16]. These models are tested sequentially and the process begins with a configural model that is well-fitting. Measurement invariance is based on how well the model fits the data as indicated by the fit indices mentioned before (i.e., CFI, RMSEA, SRMR). Additionally, change fit statistics are used to determine whether measurement invariance is present or not. This change refers to how the fit indices increase or decrease as more constraints are added. As per current recommendations, decreases below 0.01 in CFI (∆CFI) and increases below 0.015 for RMSEA (∆RMSEA) and below 0.03 for SRMR (∆SRMR) are deemed acceptable [17]. If measurement invariance was not achieved at some point, modification indices were assessed to determine whether model fit could be improved or partial invariance could be established [18]. Partial invariance occurs when some items that are different across groups are estimated freely, while keeping at least two indicators per latent construct to be equal across groups [9, 10, 19]. For the GES, this meant that only one item per dimension could be freed to attempt to establish partial invariance. If items were released, this was done following a backward method, based on the items with the highest modification indices [10]. Lastly, if no configural, metric or scalar invariance could be achieved, the factor structure was assessed separately in each group [18].\\nInternal consistency was assessed by calculating the Cronbach’s alpha coefficient [20]. Coefficients were calculated for each dimension and for the overall scale. Besides Cronbach’s alpha coefficients, composite reliability values were calculated as an additional method to evaluate the inter-item consistency of the scale. The Cronbach’s alpha value as well as the composite reliability values should be above 0.70 to be considered acceptable [21]. Floor and ceiling effects were calculated as a way to assess issues regarding responsiveness. This is an indirect way to evaluate a scale’s sensitivity to detect change [22]. Floor and ceiling effects were considered present if more than 15% of the participants achieved the lowest (i.e. 15) or highest score (i.e. 75) [22].\\nStatistical analyses were performed with the Lavaan package in R [23] and IBM SPSS Statistics for Windows version 27.\",\n \"APPROACH-IS II has its coordinating center at KU Leuven, Belgium. Therefore, ethical approval was granted from the Institutional Review Board of the University Hospitals Leuven/KU Leuven. Additionally, ethics approval was granted by the local ethics committees of the included centers (i.e., Norway and South Korea). All participants included in this study provided verbal and written informed consent.\",\n \"[SUBTITLE] Content validity [SUBSECTION] The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\\nThe proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\\n[SUBTITLE] Factorial validity [SUBSECTION] The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\\n\\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\\nFirst-order factor\\n\\nSecond-order factor\\n\\nKnowledge and Understanding\\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\\nPersonal Control\\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\\nIdentity\\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\\nShared decision-making\\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\\nEnabling Others\\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\\n\\nProportion of missing values and factor loadings for GES\\nThe five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\\n\\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\\nFirst-order factor\\n\\nSecond-order factor\\n\\nKnowledge and Understanding\\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\\nPersonal Control\\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\\nIdentity\\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\\nShared decision-making\\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\\nEnabling Others\\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\\n\\nProportion of missing values and factor loadings for GES\\n[SUBTITLE] Measurement invariance [SUBSECTION] The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\nThe configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n[SUBTITLE] Internal consistency [SUBSECTION] The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\\nThe Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\\n[SUBTITLE] Responsiveness [SUBSECTION] The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\\n\\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\\nCronbach’s alpha values\\n\\nKnowledge and understanding\\n0.7300.7210.7590.728\\nPersonal control\\n0.6930.6330.6360.751\\nIdentity\\n0.6800.6820.7510.636\\nShared decision-making\\n0.7740.8330.8040.491\\nEnabling others\\n0.8630.8720.9050.782\\nOverall scale\\n0.8730.8780.8710.854\\nComposite reliability\\n\\nKnowledge and understanding\\n0.7530.7520.7790.732\\nPersonal control\\n0.7020.6500.6770.753\\nIdentity\\n0.6840.6820.7700.657\\nShared decision-making\\n0.5590.8350.8050.470\\nEnabling others\\n0.8640.8760.9100.782\\nOverall scale\\n0.9290.9420.9510.917\\nMean patient empowerment scores\\n\\nKnowledge and understanding\\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\\nPersonal control\\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\\nIdentity\\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\\nShared decision-making\\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\\nEnabling others\\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\\nOverall scale\\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\\n\\nCronbach’s alpha values, composite reliability and mean empowerment scores\\nThe mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\\n\\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\\nCronbach’s alpha values\\n\\nKnowledge and understanding\\n0.7300.7210.7590.728\\nPersonal control\\n0.6930.6330.6360.751\\nIdentity\\n0.6800.6820.7510.636\\nShared decision-making\\n0.7740.8330.8040.491\\nEnabling others\\n0.8630.8720.9050.782\\nOverall scale\\n0.8730.8780.8710.854\\nComposite reliability\\n\\nKnowledge and understanding\\n0.7530.7520.7790.732\\nPersonal control\\n0.7020.6500.6770.753\\nIdentity\\n0.6840.6820.7700.657\\nShared decision-making\\n0.5590.8350.8050.470\\nEnabling others\\n0.8640.8760.9100.782\\nOverall scale\\n0.9290.9420.9510.917\\nMean patient empowerment scores\\n\\nKnowledge and understanding\\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\\nPersonal control\\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\\nIdentity\\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\\nShared decision-making\\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\\nEnabling others\\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\\nOverall scale\\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\\n\\nCronbach’s alpha values, composite reliability and mean empowerment scores\",\n \"The proportion of missing values ranged from 0 to 0.8% (Table 2). “Knowledge and understanding” was the dimension with the highest proportion of missing values.\",\n \"The five-factor structure as well as the overall factor of patient empowerment of the GES were evaluated through CFA in the entire sample. The five-factor model had an acceptable model fit based on the fit indices (x2(80) = 326.296; CFI = 0.948; RMSEA = 0.060; and SRMR = 0.039). Factor loadings for this model ranged from 0.515 to 0.864 and were all significant with p < 0.001 (Table 2). Item 6 from the “personal control” dimension had the lowest factor loading (i.e., 0.515). A second-order factor model to test the overall construct of patient empowerment also showed an acceptable model fit ( x2(85) = 358.916; CFI = 0.942; RMSEA = 0.062; and SRMR = 0.044). The first-order factor loadings in this model ranged from 0.523 to 0.863 and all were significant (p < 0.001). The second-order factor loadings had the following values: 0.903 (knowledge and understanding), 0.859 (personal control), 0.638 (identity), 0.773 (shared decision-making) and 0.583 (enabling others).\\n\\nTable 2Proportion of missing values and factor loadings for GESItemsMissing valuesn* (%)Factor loadings\\nFirst-order factor\\n\\nSecond-order factor\\n\\nKnowledge and Understanding\\n1. I know and understand my medical condition7 (0.8)0.7180.7232. I know what to do to stay healthy3 (0.4)0.7750.7653. I know when to contact healthcare providers for my medical condition4 (0.5)0.6320.639\\nPersonal Control\\n4. I have the skills to manage my medical condition in daily life3 (0.4)0.7790.7735. I have a sense of control over my health6 (0.7)0.6820.6836. I am active in maintaining my health0 (0)0.5150.523\\nIdentity\\n7. My medical condition is a part of who I am as a person1 (0.1)0.6920.7018. Living with my medical condition makes me stronger as a person3 (0.4)0.6100.5959. I have given my medical condition a place in my life5 (0.6)0.6420.646\\nShared decision-making\\n10. I am capable of expressing to my healthcare providers what is important to me1 (0.1)0.7150.71311. I actively participate in discussions with my health care providers about my health4 (0.5)0.7200.71912. I am capable of making decisions about my health and health care with the healthcare providers2 (0.2)0.7600.762\\nEnabling Others\\n13. I have the skills to support other people with a similar medical condition3 (0.4)0.8640.86314. I am able to give helpful advice to people who are struggling with a similar medical condition5 (0.6)0.8530.85315. I can help other people by sharing how I keep myself well1 (0.1)0.7560.757\\n\\nProportion of missing values and factor loadings for GES\",\n \"The configural model had a CFI (0.899) and RMSEA (0.085) near the cut-off values for an acceptable model (Table 3). To improve model fit, modification indices were evaluated, and based on this, the residuals of items 6 and 15 were allowed to covary. Even though these items belong to different dimensions (personal control and enabling others, respectively), it is reasonable to expect that persons who are actively involved in their care, also feel more capable of sharing their experiences with others [7]. A second configural model was evaluated with this error correlation, and model fit indices reached the expected threshold (x2(252) = 695.911; CFI = 0.913; RMSEA = 0.079; and SRMR = 0.061). By achieving a well-fitting configural model, we proceeded to test metric and scalar invariance, also including this covariation.\\nThe metric model fitted the data well (Table 3). Additionally, changes in model fit indices (∆CFI 0.005; ∆RMSEA 0.002; ∆SRMR 0.006) were within the expected values, indicating metric invariance was supported. An evaluation of scalar invariance came along with slightly worse model fit indices (Table 3). In comparison to the metric model, fit indices were not within the allowed change range (∆CFI 0.034; ∆RMSEA 0.010; ∆SRMR 0.007). Therefore, full scalar invariance could not be established.\\nGiven the lack of full scalar invariance, we proceeded to test partial scalar invariance by evaluating modification indices and the equality constraints across groups. Constraints were relaxed sequentially until an acceptable fit was achieved. Fixing the intercept of item 9 (identity) contributed the most to the observed misfit across groups, so this item was estimated freely in a partial scalar invariance model. While model fit improvements were obtained, the indices were below the acceptable threshold. Therefore, a new partial scalar invariance model was tested with the intercepts of items 9 (identity) and 4 (personal control) set free. While improvements were identified (Table 3), model fit indices were still not within an acceptable range. We continued to release items sequentially to achieve acceptable model fit. Models were tested with intercepts of items 9 (identity), 4 (personal control), 12 (shared decision-making) and 2 (knowledge and understanding) unconstrained and while fit indices were almost within the recommended range, it was not possible to achieve acceptable model fit. While models with unconstrained items from the “enabling others” dimension were assessed, none of them led to changes in the model fit indices. Additionally, given that not more than one item per dimension could be estimated freely, models with more than 5 free items where not tested. Therefore, partial scalar invariance was rejected.\\nSince partial scalar invariance was not achieved, an assessment of the three groups independently through CFA was undertaken to understand why this was the case. While fit indices for Norway (x2(85) = 117.694; CFI = 0.964; RMSEA = 0.053; and SRMR = 0.074) and Belgium ( x2(85) = 348.280; CFI = 0.916; RMSEA = 0.079; and SRMR = 0.053) indicated the model fitted these groups well, it appears this was not the case for the data from South Korea (x2(85): 307.753; CFI: 0.788; RMSEA: 0.112; and SRMR: 0.079). Factor loadings for this country ranged between 0.407 and 0.772, with low factor loadings in the “shared decision-making” dimension, though significant (p < 0.001).\\n[SUBTITLE] Measurement invariance: norwegian and belgian samples [SUBSECTION] As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\nAs the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\",\n \"As the CFA models for Norway and Belgium had an acceptable model fit. An evaluation of measurement invariance within these two countries was undertaken. Configural invariance was supported by a well-fitted model (x2(170) = 463.510; CFI = 0.928; RMSEA = 0.074; and SRMR = 0.057). An evaluation of metric invariance also showed acceptable model fit indices (Table 3) and changes in them were also within the acceptable ranges (∆CFI 0.001; ∆RMSEA 0.002; ∆SRMR 0.004). Given that metric invariance was supported, scalar invariance was evaluated next. Fit indexes for this model were within the recommended values (Table 3). However, changes in the CFI were outside of the acceptable range to support measurement invariance (∆CFI 0.013; ∆RMSEA 0.004; ∆SRMR 0.002). Therefore, full scalar invariance was not supported. Modification indices indicated that freeing the intercept of item 11 (shared decision-making) could lead to model improvements. Hence, partial scalar invariance was evaluated with the intercept of item 11 free. This model had acceptable model fit indexes (x2(192) = 519.18; CFI = 0.920; RMSEA = 0.073; and SRMR = 0.063). Additionally, in comparison to the metric model, fit indices suggested worse model fit, but changes were still within acceptable ranges (∆CFI 0.007; ∆RMSEA 0.002; ∆SRMR 0.001). These results support partial scalar invariance for the samples of Norway and Belgium.\\n\\nTable 3Model fit indexes of multi-group confirmatory factor analysesModelX2 (df)CFIRMSEASRMR∆X2 (∆df)∆CFI∆RMSEA∆SRMR\\nComplete sample (Belgium, Norway and South Korea)\\n\\nConfigural invariance\\na\\n772.074 (255)0.8990.0850.063\\nConfigural invariance\\nb\\n695.911 (252)0.9130.0790.061\\nMetric invariance\\nb\\n750.492 (280)0.9080.0770.06754.581 (28)0.0050.0020.006\\nScalar invariance\\nb\\n941.720 (298)0.8740.0870.074191.228 (18)0.0340.0100.007\\nPartial scalar invariance with intercept of item 9 free\\nb\\n886.943 (296)0.8850.0840.072\\nPartial scalar invariance with intercept of items 9 and 4 free\\nb\\n856.568 (294)0.8900.0820.070\\nPartial scalar invariance with intercept of items 9, 4 and 12 free\\nb\\n827.343 (292)0.8960.0800.069\\nPartial scalar invariance with intercept of items 9, 4, 12 and 2 free\\nb\\n813.205 (290)0.8980.0800.069\\nPartial sample (Belgium and Norway)\\n\\nConfigural invariance\\na\\n463.510 (170)0.9280.0740.057\\nMetric invariance\\na\\n482.543 (184)0.9270.0710.06119.032 (14)0.0010.0020.004\\nScalar invariance\\na\\n544.309 (193)0.9140.0750.06461.767 (9)0.0130.0040.002\\nPartial scalar invariance with one intercept free\\nc\\n519.177 (192)0.9200.0730.06336.635 (8)0.0070.0020.001a Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\\n\\nModel fit indexes of multi-group confirmatory factor analyses\\na Model without error correlation; b Model with error correlation between items 6 and 15. Models were considered to have acceptable fit if CFI > 0.90, RMSEA and SRMR values < 0.08. MGCFA needed also to have it indexes changes within these ranges: ΔCFI < 0.010, ΔRMSEA < 0.015, and ΔSRMR < 0.030\",\n \"The Cronbach’s alpha for the overall scale was 0.873, indicating that it was internally consistent. The alpha values for the subscales were: 0.73 (knowledge and understanding); 0.693 (personal control); 0.680 (identity); 0. 774 (shared decision-making); and 0.863 (enabling others). The Cronbach’s alpha for each country is given in Table 4. An evaluation of the scale’s composite reliability showed that most of the values are above the expected range of 0.7 and those below such threshold were relatively near to this acceptable value.\",\n \"The mean patient empowerment score for the entire sample was 59.36 ± 8.47. None of the participants had the lowest score (i.e., 15) and only 2.5% had the highest attainable score of 75. Hence, no floor or ceiling effects were identified.\\n\\nTable 4Cronbach’s alpha values, composite reliability and mean empowerment scoresTotal SampleBelgiumNorwaySouth Korea\\nCronbach’s alpha values\\n\\nKnowledge and understanding\\n0.7300.7210.7590.728\\nPersonal control\\n0.6930.6330.6360.751\\nIdentity\\n0.6800.6820.7510.636\\nShared decision-making\\n0.7740.8330.8040.491\\nEnabling others\\n0.8630.8720.9050.782\\nOverall scale\\n0.8730.8780.8710.854\\nComposite reliability\\n\\nKnowledge and understanding\\n0.7530.7520.7790.732\\nPersonal control\\n0.7020.6500.6770.753\\nIdentity\\n0.6840.6820.7700.657\\nShared decision-making\\n0.5590.8350.8050.470\\nEnabling others\\n0.8640.8760.9100.782\\nOverall scale\\n0.9290.9420.9510.917\\nMean patient empowerment scores\\n\\nKnowledge and understanding\\n12.94 (1.90)13.07 (1.94)12.93 (2.03)12.61 (1.68)\\nPersonal control\\n12.04 (2.10)12.36 (1.97)12.59 (1.85)10.90 (2.12)\\nIdentity\\n11.66 (2.39)11.82 (2.42)11.19 (2.79)11.58 (1.97)\\nShared decision-making\\n12.15 (2.26)12.55 (2.22)11.58 (2.75)11.61 (1.74)\\nEnabling others\\n10.57 (2.83)10.72 (2.90)10.51 (3.12)10.25 (2.40)\\nOverall scale\\n59.37 (8.47)60.52 (8.53)58.86 (9.07)56.98 (7.33)\\n\\nCronbach’s alpha values, composite reliability and mean empowerment scores\",\n \"Recent research highlights the need to improve and evaluate patient empowerment in persons with CCs [4]. However, the availability of instruments with a strong theoretical background and/or acceptable psychometric properties is limited [6]. Therefore, the present study evaluated a slightly modified version of a previously validated patient empowerment in a group of adults with CHD. Results indicate that GES proved reliable, meaning the scale’s items altogether are consistent. Additionally, no floor or ceiling effects were identified, which is indirectly related to the scale’s sensitivity to measure change [22]. Whereas these results are in favor of the GES, a psychometric assessment of the scale’s validity revealed that the scale appears valid in two of the three included countries.\\nResults from the CFA of the entire sample indicate the five-factor structure fits the data well. However, due to the lack of measurement invariance, CFAs were performed separately for patients from each country. Analyses revealed that an acceptable model fit was only achieved for the samples from Belgium and Norway. For the South Korean data, the CFI and the RMSEA were above the acceptable ranges although factor loadings from the “shared decision-making” dimension were low. This is an indication that the relations among the latent constructs of GES are not in line with the data, i.e., the data do not support the model [24]. Even though it is common in CFA to conduct an evaluation of modification indexes to improve model fit, changes should be theoretically justified [25]. In this case, we could identify no strong theoretical reasons to allow for model changes that could potentially improve the model fit.\\nThe poorly fitted model to the South Korean data might be the reason why invariance was not achieved when comparing the three countries. A lack of invariance suggests that patient empowerment may have a different meaning in different countries. Therefore, making mean comparisons of this construct across countries using the GES should be made with caution. Nonetheless, while cross-country comparisons cannot be made, it is possible to compare how the GES relates to other constructs (i.e., variables) within countries. For example, one could undertake a study to compare how the GES relates to patient functioning in Belgium, vs. how the GES relates to patient functioning in South Korea. Another potential research aim would be to determine whether changes in GES relate to psychological distress and whether these changes have the same magnitude in Belgium, Norway and South Korea.\\nThe absence of invariance across the three countries can be an indicator of patient empowerment being interpreted differently across countries. Perhaps individuals in South Korea have different interpretations and values of patient empowerment compared to patients in Belgium and Norway, two European countries that are likely more culturally similar. It has been suggested that patient empowerment is determined by its context and that individuals within a particular environment, organization or country may have a different perception of empowerment [26]. Hence, a measure that fits all persons (or contexts), might be hard to develop. Findings from the present study partially support this notion. It is plausible to consider that different countries (and cultures) have a different understanding of patient empowerment domains, values, and skills.\\nEvidence on how patient empowerment may differ between countries is limited. However, authors of a systematic review concluded that although patients do share many perspectives of this construct, there might be variations on how certain aspects of patient empowerment are understood and research has yet to address structural elements of patient empowerment [27].\\nConsideration of structural elements is relevant for the current study because the “shared decision-making” dimension had the lowest factor loadings for the South Korean sample, as well as poor reliability. Shared decision-making involves healthcare professionals and patients working together on a care plan [28]. It entails patients who are willing to participate as well as clinicians who want to collaborate with the patient. Hence, it is greatly influenced by the healthcare structure. Within Asian cultures, for example, it is more commonly believed that patients prefer to take a less independent role and that a family-centered (rather than individual-centered) approach towards decision-making is preferred [29–31]. It has been suggested that shared decision-making reflects values associated with western cultures and that this might be different in other cultures [28, 31, 32]. The value placed on shared decision-making and the structure of the healthcare system might therefore potentially impact the value of patient empowerment as perceived by study participants. Therefore, the comparison of this construct between different cultures and regions of the world should be made with caution.\\nAlthough the GES appears to be a valid and reliable instrument for the Norwegian and Belgian sample, it is worth noting that future research should evaluate whether the low psychometric properties found in the South Korean sample are indeed applicable to other Asian or non-European countries. Comparatives studies evaluating this construct (and similar constructs) between Western and Asian countries are needed to comprehend this phenomenon, the types of research questions that can be addressed and the conclusions that can be drawn when evaluating outcomes such as patient empowerment. Such studies could be undertaken once the data collection from APPROACH-IS II is finalized.\\n[SUBTITLE] Methodological considerations [SUBSECTION] There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.\\nThere are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.\",\n \"There are several strengths of the study that merit mentioning. First, GES development was based on a previously validated questionnaire, which in turn has a strong theoretical foundation and followed a rigorous development process. Therefore, it is plausible to conclude that GES also has a strong theoretical foundation. Second, data were collected as part of an international study, which allows for cross-country comparisons. Third, the data were collected before the COVID-19 pandemic, therefore there was no concern about the potential for survey responses to be impacted by the pandemic [33].\\nThere are, however, also some methodological limitations associated with this study. First, some aspects associated with validity and reliability were not assessed. For instance, it was not possible to directly evaluate responsiveness with cross-sectional data. While floor and ceiling effects are an indirect way to measure responsiveness, this is better achieved through longitudinal studies. Future longitudinal studies should evaluate this psychometric property. Second, the study only includes individuals with CHD, who were in current follow-up and who had the abilities to answer the questionnaires. Therefore, the generalizability of the results might be limited.\",\n \"This study provides evidence on the GES’s validity, reliability, and responsiveness in adults with CHD. GES seems to be valid and reliable for the sample of Belgium and Norway. However, this is not the case for South Korea, because shared decision-making seems to have a different meaning in Asian cultures. Hence, cross-cultural comparisons using GES should be made cautiously. It is possible that in countries who are more culturally different from Norway or Belgium, the scale might not perform as well.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,"results",null,null,null,null,null,null,"discussion",null,"conclusion"],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n \"conclusion\"\n]"},"keywords":{"kind":"list like","value":["Adults","Congenital heart defects","Chronic conditions","Measurement invariance","Patient empowerment","Psychometrics","Validity","Reliability"],"string":"[\n \"Adults\",\n \"Congenital heart defects\",\n \"Chronic conditions\",\n \"Measurement invariance\",\n \"Patient empowerment\",\n \"Psychometrics\",\n \"Validity\",\n \"Reliability\"\n]"}}},{"rowIdx":371,"cells":{"title":{"kind":"string","value":"Detection of C8/T1 radiculopathy by measuring the root motor conduction time."},"pmid":{"kind":"string","value":"36266617"},"background_abstract":{"kind":"string","value":"Root motor conduction time (RMCT) can noninvasively evaluate the status of the proximal root segment. However, its clinical application remains limited, and wider studies regarding its use are scarce. We aimed to investigate the association between C8/T1 level radiculopathy and RMCT."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"This was a retrospective cross-sectional study. Subjects were extracted from a general hospital's spine clinic database. A total of 48 C8/T1 root lesions from 37 patients were included, and 48 C8/T1 root levels from control subjects were matched for age, sex, and height. RMCT was measured in the abductor pollicis brevis muscle and the assessment of any delays owing to C8/T1 radiculopathy."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The RMCT of the C8/T1 radiculopathy group was 1.7 ± 0.6 ms, which was significantly longer than that in the control group (1.2 ± 0.8 ms; p = 0.001). The delayed RMCT was independently associated with radiculopathy (adjusted odds ratio, 1.15; 95% confidence interval, 1.06-1.27; p = 0.011) after adjusting for the peripheral motor conduction time, amplitude of median compound motor nerve action potential, and shortest F-wave latency. The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61-0.82). The RMCT was significantly correlated with symptom duration (coefficient = 0.58; p < 0.001) but was not associated with the degree of arm pain."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Our findings illustrate the clinical applicability of the RMCT by demonstrating its utility in diagnosing radiculopathy at certain spinal levels."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Radiculopathy","Retrospective Studies","Cross-Sectional Studies","Muscle, Skeletal","Evoked Potentials","Neural Conduction"],"string":"[\n \"Humans\",\n \"Radiculopathy\",\n \"Retrospective Studies\",\n \"Cross-Sectional Studies\",\n \"Muscle, Skeletal\",\n \"Evoked Potentials\",\n \"Neural Conduction\"\n]"},"pmcid":{"kind":"string","value":"9583482"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Radiculopathy is a common condition whose symptoms can include pain, sensory change, and motor weakness owing to mechanical and chemical irritation of the spinal nerve root [1, 2]. Imaging studies are essential for diagnosing radiculopathy; magnetic resonance imaging helps identify soft tissues such as discs, ligaments, and nerve roots [3], while computed tomography is primarily used to confirm abnormalities in bony structures [4]. Meanwhile, another key factor in confirming radiculopathy is electrophysiologic diagnosis [5]. In particular, electromyography (EMG) is a classical tool that was used to evaluate radiculopathy before the development of modern imaging instruments [6]; its key advantage is in its ability to detect neurophysiological changes that cannot be confirmed by imaging studies alone [7]. By applying EMG to multiple myotomes, it is possible not only to localize the abnormal spinal root level but also to estimate the temporal aspect of compression by distinguishing motor unit action potential [8]. However, the invasive nature of EMG creates a potential risk of complications such as bleeding or infection at the examination site [9]. Additionally, patient cooperation is essential for observing resting potential and interference patterns [10].\nRoot motor conduction time (RMCT) is a measure of the state of the proximal root segment that is obtained noninvasively by performing direct cervical stimulation and nerve conduction studies [11, 12]. Notably, RMCT is very localized since it excludes distal neural lesions and measures only the conduction time of the proximal root segment [13–15]. However, it also has the disadvantage of involving a somewhat complicated measurement method wherein well-trained examiners are required [16]. To date, the RMCT has been investigated in only a few disease types, such as demyelinating disease and lumbar spinal stenosis, whereas its clinical applicability in patients with radiculopathy remains unclear [17].\nIn this study, we hypothesized that the RMCT can be a valuable tool for the diagnosis of radiculopathy at certain spinal root levels. To that end, we investigated whether the RMCT measured in the abductor pollicis brevis (APB) muscle was significantly delayed in the presence of C8/T1 radiculopathy. Moreover, we aimed to determine the association between clinical symptoms and RMCT."},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Baseline characteristics of patients and controls [SUBSECTION] Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\nTable 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n[SUBTITLE] RMCT at the C8/T1 level [SUBSECTION] The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\nThe mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"We demonstrated that the RMCT is delayed in the root lesions of patients with C8/T1 level radiculopathy. This noninvasive method (compared to EMG) can have an adjuvant role in diagnosing radiculopathy at certain spinal levels. Our study may be a good milestone for future clinical applications of RMCT as related to patients with radiculopathy. However, it is important to validate our results through additional multi-center, multi-ethnic studies."},"other_sections_titles":{"kind":"list like","value":["Methods","Subjects and clinical evaluations","Electrodiagnostic assessments","Statistical analysis","Baseline characteristics of patients and controls","RMCT at the C8/T1 level",""],"string":"[\n \"Methods\",\n \"Subjects and clinical evaluations\",\n \"Electrodiagnostic assessments\",\n \"Statistical analysis\",\n \"Baseline characteristics of patients and controls\",\n \"RMCT at the C8/T1 level\",\n \"\"\n]"},"other_sections_texts":{"kind":"list like","value":["[SUBTITLE] Subjects and clinical evaluations [SUBSECTION] We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\nWe performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\n[SUBTITLE] Electrodiagnostic assessments [SUBSECTION] The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\nThe EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\n[SUBTITLE] Statistical analysis [SUBSECTION] The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\nThe Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.","We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.","The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.","The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.","Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time","The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"[SUBTITLE] Subjects and clinical evaluations [SUBSECTION] We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\\nWe performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\\n[SUBTITLE] Electrodiagnostic assessments [SUBSECTION] The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\\nThe EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\\n[SUBTITLE] Statistical analysis [SUBSECTION] The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\\nThe Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\",\n \"We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\",\n \"The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\",\n \"The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\",\n \"Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\\n\\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\n\\nBaseline features of patients with C8/T1 level radiculopathy\\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\\n\\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\n\\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\",\n \"The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\\n\\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\\n\\nLogistic regression models using data from patients and matched control subjects\\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Subjects and clinical evaluations","Electrodiagnostic assessments","Statistical analysis","Results","Baseline characteristics of patients and controls","RMCT at the C8/T1 level","Discussion","Conclusion","Electronic supplementary material",""],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Subjects and clinical evaluations\",\n \"Electrodiagnostic assessments\",\n \"Statistical analysis\",\n \"Results\",\n \"Baseline characteristics of patients and controls\",\n \"RMCT at the C8/T1 level\",\n \"Discussion\",\n \"Conclusion\",\n \"Electronic supplementary material\",\n \"\"\n]"},"all_sections_texts":{"kind":"list like","value":["Radiculopathy is a common condition whose symptoms can include pain, sensory change, and motor weakness owing to mechanical and chemical irritation of the spinal nerve root [1, 2]. Imaging studies are essential for diagnosing radiculopathy; magnetic resonance imaging helps identify soft tissues such as discs, ligaments, and nerve roots [3], while computed tomography is primarily used to confirm abnormalities in bony structures [4]. Meanwhile, another key factor in confirming radiculopathy is electrophysiologic diagnosis [5]. In particular, electromyography (EMG) is a classical tool that was used to evaluate radiculopathy before the development of modern imaging instruments [6]; its key advantage is in its ability to detect neurophysiological changes that cannot be confirmed by imaging studies alone [7]. By applying EMG to multiple myotomes, it is possible not only to localize the abnormal spinal root level but also to estimate the temporal aspect of compression by distinguishing motor unit action potential [8]. However, the invasive nature of EMG creates a potential risk of complications such as bleeding or infection at the examination site [9]. Additionally, patient cooperation is essential for observing resting potential and interference patterns [10].\nRoot motor conduction time (RMCT) is a measure of the state of the proximal root segment that is obtained noninvasively by performing direct cervical stimulation and nerve conduction studies [11, 12]. Notably, RMCT is very localized since it excludes distal neural lesions and measures only the conduction time of the proximal root segment [13–15]. However, it also has the disadvantage of involving a somewhat complicated measurement method wherein well-trained examiners are required [16]. To date, the RMCT has been investigated in only a few disease types, such as demyelinating disease and lumbar spinal stenosis, whereas its clinical applicability in patients with radiculopathy remains unclear [17].\nIn this study, we hypothesized that the RMCT can be a valuable tool for the diagnosis of radiculopathy at certain spinal root levels. To that end, we investigated whether the RMCT measured in the abductor pollicis brevis (APB) muscle was significantly delayed in the presence of C8/T1 radiculopathy. Moreover, we aimed to determine the association between clinical symptoms and RMCT.","[SUBTITLE] Subjects and clinical evaluations [SUBSECTION] We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\nWe performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\n[SUBTITLE] Electrodiagnostic assessments [SUBSECTION] The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\nThe EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\n[SUBTITLE] Statistical analysis [SUBSECTION] The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\nThe Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.","We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\n\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\n\nFlowchart showing the comparative analysis of the patients and control subjects\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.","The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\n\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\n\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\n\n\nRMCT (ms) = calculated PMCT − stimulated PMCT\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.","The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.","[SUBTITLE] Baseline characteristics of patients and controls [SUBSECTION] Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\nTable 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n[SUBTITLE] RMCT at the C8/T1 level [SUBSECTION] The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\nThe mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency","Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\n\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\n\nBaseline features of patients with C8/T1 level radiculopathy\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\n\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\n\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time","The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\n\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\n\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\n\nLogistic regression models using data from patients and matched control subjects\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency","We observed a significant delay in RMCT among patients with C8/T1 radiculopathy when compared to control subjects, and also demonstrated that this delay was independently associated with the presence of radiculopathy at the C8/T1 level. This finding is notable because it provided a rational basis for the applicability of RMCT in terms of diagnosing radiculopathy at a specific spinal root level.\nThe application of RMCT to assess radiculopathy occurring at particular levels is rare. Most relevant studies were conducted before the 2000s, with only a few published since then. Banerjee et al. [13] described magnetic spinal stimulation as a non-invasive method to evaluate lumbosacral motor radiculopathy in 26 patients and 25 control subjects; they mainly targeted the lower lumbosacral nerve roots (L5, S1, and S2) by recording from the abductor hallucis muscle. According to their results, patients with clinical motor weakness had profoundly prolonged RMCTs; therefore, they suggested that RMCT delay was related to the severity of symptoms. Golez [22] measured motor conduction time via stimulation at the L1 and S1 levels in 25 patients with lumbar spinal stenosis and 36 control subjects. That study was noteworthy because it demonstrated that neurogenic claudication increased the cauda equina motor conduction time, revealing that the clinical symptom and delay of motor conduction time were related. In our study, the RMCT showed a significant correlation with the duration of subjective symptoms, which supports Golez’s results to some extent. However, in our study, the subjective pain level in the lesional-side limb was not related to RMCT as we assumed when we primarily evaluated the motor nerve. Additionally, Seçil et al. [23] measured the cauda equina motor conduction time and showed that it was slower in the lumbar spinal stenosis group than in the control group; however, their study did not identify the relationship between the degree of motor conduction delay and symptom severity.\nOur study is the first to measure RMCT in patients with cervical radiculopathy and has the advantage of being specific to a specific spinal level; moreover, we included a relatively large number of subjects compared to previous related studies. Additionally, variables such as age, sex, and height were analyzed and adjusted for using PSM, rendering the results much more reliable. Since the APB-RMCT was previously found to be affected by height in a linear model [18], for the integrity of our results, it was important that we controlled for this factor when assessing the patient and control groups.\nRMCT has also been studied in patients with demyelinating neuropathy. In a previous study, a significant difference was found in the RMCT of 30 healthy subjects and 12 patients with Guillain-Barre syndrome who were diagnosed within one week [14]; hence, RMCT was deemed to be useful for the early diagnosis of focal segmental demyelinating polyneuropathy, affording this measure a novel clinical utility. In the study of Inaba et al. [24], RMCT was measured in 11 patients with chronic inflammatory demyelinating neuropathy and 10 with Guillain-Barre syndrome, wherein the authors found that the RMCT increasingly normalized as muscle strength recovered. Taken together, such studies illustrate the applicability of the RMCT to various diseases involving the motor root segment. Furthermore, measuring the RMCT is a non-invasive procedure and has the potential to be applied to additional fields given its ability to detect diseases in their early stages and to also assess functional recovery. Therefore, additional research on the use of the RMCT is warranted going forward.\nThis study had several limitations. It was retrospective in nature, requiring additional investigations to validate our results. Additionally, our participants were limited to Koreans; hence, validation among patients of other ethnicities is also required. Our patients comprised a single group diagnosed with C8/T1 radiculopathy that was relatively heterogeneous; future studies that analyze various subgroups of patients (such as those with acute versus chronic lesions or mild versus severe symptoms), would further clarify the relationship between RMCT and radiculopathy. Finally, there were potential limitations in terms of measurement. For example, it is difficult to apply our methods at some spinal root levels because reliable recording of the F-wave is possible only from distal muscles [25]. Moreover, it was challenging to measure the RMCT when reliable motor evoked potentials or compound muscle action potentials could not be obtained owing to severe peripheral nerve lesions in the distal limbs.","We demonstrated that the RMCT is delayed in the root lesions of patients with C8/T1 level radiculopathy. This noninvasive method (compared to EMG) can have an adjuvant role in diagnosing radiculopathy at certain spinal levels. Our study may be a good milestone for future clinical applications of RMCT as related to patients with radiculopathy. However, it is important to validate our results through additional multi-center, multi-ethnic studies.","[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\nBelow is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"Radiculopathy is a common condition whose symptoms can include pain, sensory change, and motor weakness owing to mechanical and chemical irritation of the spinal nerve root [1, 2]. Imaging studies are essential for diagnosing radiculopathy; magnetic resonance imaging helps identify soft tissues such as discs, ligaments, and nerve roots [3], while computed tomography is primarily used to confirm abnormalities in bony structures [4]. Meanwhile, another key factor in confirming radiculopathy is electrophysiologic diagnosis [5]. In particular, electromyography (EMG) is a classical tool that was used to evaluate radiculopathy before the development of modern imaging instruments [6]; its key advantage is in its ability to detect neurophysiological changes that cannot be confirmed by imaging studies alone [7]. By applying EMG to multiple myotomes, it is possible not only to localize the abnormal spinal root level but also to estimate the temporal aspect of compression by distinguishing motor unit action potential [8]. However, the invasive nature of EMG creates a potential risk of complications such as bleeding or infection at the examination site [9]. Additionally, patient cooperation is essential for observing resting potential and interference patterns [10].\\nRoot motor conduction time (RMCT) is a measure of the state of the proximal root segment that is obtained noninvasively by performing direct cervical stimulation and nerve conduction studies [11, 12]. Notably, RMCT is very localized since it excludes distal neural lesions and measures only the conduction time of the proximal root segment [13–15]. However, it also has the disadvantage of involving a somewhat complicated measurement method wherein well-trained examiners are required [16]. To date, the RMCT has been investigated in only a few disease types, such as demyelinating disease and lumbar spinal stenosis, whereas its clinical applicability in patients with radiculopathy remains unclear [17].\\nIn this study, we hypothesized that the RMCT can be a valuable tool for the diagnosis of radiculopathy at certain spinal root levels. To that end, we investigated whether the RMCT measured in the abductor pollicis brevis (APB) muscle was significantly delayed in the presence of C8/T1 radiculopathy. Moreover, we aimed to determine the association between clinical symptoms and RMCT.\",\n \"[SUBTITLE] Subjects and clinical evaluations [SUBSECTION] We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\\nWe performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\\n[SUBTITLE] Electrodiagnostic assessments [SUBSECTION] The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\\nThe EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\\n[SUBTITLE] Statistical analysis [SUBSECTION] The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\\nThe Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\",\n \"We performed a retrospective cross-sectional study based on medical records of patients treated between June 2007 and December 2021. We extracted patients diagnosed with C8/T1 radiculopathy at our hospital through electrodiagnosis and whose RMCT was measured at the same time. All patients diagnosed with C8/T1 radiculopathy also underwent cervical x-ray and magnetic resonance imaging studies to confirm root compression and exclude other differential diagnoses. Thereafter, the final study group was selected by applying the following exclusion criteria: uncontrolled diabetes, concomitant polyneuropathy or median nerve lesion, previous cervical spine surgery, previous hand injury or surgery, unobtainable electrodiagnostic parameters, and lack of clinical information. Data from 48 C8/T1 root lesions in 37 patients were finally extracted; these included the subjective symptom duration, numerical rating scale (NRS) of the neck and arm pain, and neck disability index (NDI) score at the time of electrodiagnostic examination.\\nThe control group consisted of a single-center healthy cohort of 190 prospectively analyzed subjects who were reported previously by the authors [18]. We controlled the potential biases between the two groups by performing propensity score matching (PSM). Consequently, 96 root levels (48 patients and 48 matched controls) were finally investigated (Fig. 1).\\n\\nFig. 1Flowchart showing the comparative analysis of the patients and control subjects\\n\\nFlowchart showing the comparative analysis of the patients and control subjects\\nThis study was approved by the Institutional Review Board of Pohang Stroke and Spine Hospital (PSSH0475-202207-HR-011-01); informed consent was waived, given its retrospective design. The research complied with the guidelines of the Declaration of Helsinki.\",\n \"The EMG protocol for radiculopathy comprised of the following [5]: First, we examined muscles innervated by different nerves in the same myotome. Second, we assessed both proximal and distal muscles in the same myotome. Finally, we examined muscles in the myotomes adjacent to the suspected level. EMG-confirmed C8/T1 radiculopathy was defined as the presence of denervation potentials or polyphasic, long-duration, and large-amplitude motor unit action potentials in two or more of the differently innervated following muscles after other conditions were ruled out: flexor pollicis longus, flexor digitorum profundus, flexor carpi radialis, first dorsal interosseous, abductor digiti minimi, and APB [5].\\nFor the nerve conduction study, we first recorded the sensory nerve action potentials of the median, ulnar, and superficial radial nerve in the distal arm. Electrical stimulations with 0.1 ms square wave pulses were applied for the sensory nerve conduction studies with a 10–2000 Hz filter setting. Next, we recorded compound motor nerve action potentials from the APB and abductor digiti minimi muscles and F-waves from the APB muscle. To conduct motor nerve conduction studies, each nerve was supramaximally stimulated by 0.2 ms square wave pulses with a 5–5000 Hz filter setting. We recorded nerve action potentials at least 12 times at the same nerve for study reproducibility. Surface electrodes were attached using the belly-tendon method to all recorded muscles. All nerve conduction studies were performed in the supine position. The detailed methods for individual nerve conduction studies are described in Supplementary Table 1 [5, 19].\\n\\nTo measure the stimulated peripheral motor conduction time (PMCT), we performed magnetic stimulation at the C7 spinous process; the ABP muscle was recorded using the surface electrodes. To provoke median motor evoked potential, we administered supramaximal stimulation (20% above the threshold) with weak isometric contraction in the APB muscle. The simulations were applied biphasic and the active pulse width was 280 µs. Then, we calculated the RMCT using the following equations, which are based on the principle that nerve excitement occurs a few centimeters distal to the anterior horn cell upon magnetic stimulation (Fig. 2) [20].\\n\\nFig. 2Schematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\nSchematic diagram illustrating the measurement of C8/T1 root motor conduction time\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCTCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2Stimulated PMCT (ms) = spinal motor-evoked potential onset latency\\n\\n\\nRMCT (ms) = calculated PMCT − stimulated PMCT\\nCalculated PMCT (ms) = (compound muscle action potential onset latency + F-wave onset latency − 1) / 2\\nStimulated PMCT (ms) = spinal motor-evoked potential onset latency\\nAll electrodiagnostic evaluations were performed using the Cadwell Sierra Wave EMG system (Cadwell Laboratories Inc., Kennewick, WA, USA). A MagPro Compact with a C-100 circular coil (11 cm outer diameter) (MagVenture Inc, Farum, Denmark) was used for cervical magnetic stimulation. The temperature of the electrodiagnostic laboratory was maintained at approximately 23℃ to 25℃. All electrodiagnostic examinations were interpreted by experienced physiatrists.\",\n \"The Shapiro-Wilk test was applied to check the normality of continuous variables. These are expressed as means ± standard deviations if normality was satisfied or as medians (interquartile ranges) if not; the independent t-test and Mann-Whitney U-test were applied for comparative analyses of these types of values, respectively. Categorical variables are expressed as frequencies (proportions); groups were compared using the chi-squared test. We established binary logistic regression models for detecting C8/T1 root lesions using data adjusted for age, sex, and height. Model 1 was a predictive model of RMCT alone, Model 2 was adjusted for the stimulated PMCT, and Model 3 was adjusted for the median compound muscle action potential and F-wave in addition to the adjustment of Model 2. The multicollinearity of the model was confirmed, as the variance inflation factor was < 10. We drew the Receiver Operating Characteristic curve and calculated the cutoff value using Youden’s J statistic. We calculated the Spearman coefficient to determine the correlation between RMCT and symptoms. All statistical analyses were performed using SPSS 22.0 (IBM Inc., Armonk, NY, USA).\\nFor PSM, the “MatchIt” package of the R software version 4.1.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) was used [21]. The covariates used for matching were age, sex, and height; moreover, 1:1 matching with no replacement was performed. The nearest-neighbor method was applied with the caliper set to 0.2.\",\n \"[SUBTITLE] Baseline characteristics of patients and controls [SUBSECTION] Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\\n\\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\n\\nBaseline features of patients with C8/T1 level radiculopathy\\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\\n\\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\n\\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\nTable 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\\n\\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\n\\nBaseline features of patients with C8/T1 level radiculopathy\\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\\n\\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\n\\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\n[SUBTITLE] RMCT at the C8/T1 level [SUBSECTION] The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\\n\\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\\n\\nLogistic regression models using data from patients and matched control subjects\\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\\nThe mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\\n\\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\\n\\nLogistic regression models using data from patients and matched control subjects\\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\",\n \"Table 1 summarizes the baseline characteristics of the 37 patients, 11 of whom had bilateral C8/T1 radiculopathy. The mean patient age was 61.1 years and the mean height was 166.5 cm; 75.7% were men. The patients’ median NDI score was 12.0. Among all 48 C8/T1 root levels, 25 (52.1%) were right-sided, and the median duration of symptoms was 4.0 (3.0–12.0) months. The median NRS scores of the neck and arm were 3.0 and 4.5, respectively.\\n\\nTable 1Baseline features of patients with C8/T1 level radiculopathyVariablesValueaTotal patients, n37Age, years61.1 ± 12.2Male, n (%)28 (75.7)Height, cm166.5 ± 8.7Hypertension, n (%)16 (43.2)Diabetes, n (%)5 (13.5)Dyslipidemia, n (%)6 (16.2)NRS, neck3.0 (2.0–4.5)Total C8/T1 root levels, n48Right side, n (%)25 (52.1)Symptom duration, months4.0 (3.0–12.0)NDI12.0 (5.5–19.0)NRS, arm4.5 (3.0–6.0)Abbreviations: NDI, neck disability index; NRS, numerical rating scaleaUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\n\\nBaseline features of patients with C8/T1 level radiculopathy\\nAbbreviations: NDI, neck disability index; NRS, numerical rating scale\\naUnless otherwise indicated, values are means ± standard deviations or medians (interquartile ranges)\\nAfter performing PSM for comparative analysis, we extracted 48 control C8/T1 root levels and confirmed that there were no significant differences in age, sex, or height between the two groups (Table 2).\\n\\nTable 2Comparison of data from patients with C8/T1 level radiculopathy and matched control subjectsPatients (n = 48)Controls (n = 48)p-valueAge, years61.9 ± 12.458.8 ± 11.10.197Male, n (%)37 (77.1)28 (58.3)0.081Height, cm166.4 ± 8.6164.2 ± 8.80.223APB-RMCT, ms1.7 ± 0.61.2 ± 0.80.001Abbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\\n\\nComparison of data from patients with C8/T1 level radiculopathy and matched control subjects\\nAbbreviations: APB, abductor pollicis brevis; RMCT, root motor conduction time\",\n \"The mean RMCT among patients with radiculopathy was 1.7 ± 0.6 ms, which was significantly longer than that among control subjects (1.2 ± 0.8 ms; p = 0.001) (Table 2; Fig. 3). Logistic regression models revealed that the delay in RMCT observed in patients with C8/T1 was independently associated with radiculopathy (per Model 3: odds ratio, 1.15; 95% confidence interval, 1.06–1.27; p = 0.011) (Table 3). The area under the Receiver Operating Characteristic curve for diagnosing C8/T1 radiculopathy using RMCT was 0.72 (0.61–0.82) with 0.83 sensitivity and 0.58 specificity (Supplementary Fig. 1). In terms of subjective symptom indices, the RMCT was significantly associated with symptom duration (coefficient = 0.58; p < 0.001) but not with the NRS of the arm (coefficient = 0.22; p = 0.139) All measured values for calculating the RMCT are presented in Supplementary Table 2.\\n\\nFig. 3Root motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nRoot motor conduction time (RMCT) in each group. Patients with C8/T1 level radiculopathy showed significantly longer RMCTs than did the control group (p = 0.001)\\n\\nTable 3Logistic regression models using data from patients and matched control subjectsOdds ratio (95% CI)p-valueModel 1aRMCT (per 0.1 ms)1.11 (1.04–1.18)0.002Model 2bRMCT (per 0.1 ms)1.16 (1.06–1.27)0.002Model 3cRMCT (per 0.1 ms)1.15 (1.06–1.27)0.011Abbreviations: CI, confidence interval; RMCT, root motor conduction timeaUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\\n\\nLogistic regression models using data from patients and matched control subjects\\nAbbreviations: CI, confidence interval; RMCT, root motor conduction time\\naUnivariable analysis; badjusted for stimulated peripheral motor conduction time; cadjusted for stimulated peripheral motor conduction time, amplitude of median compound motor nerve action potential, and F-wave latency\",\n \"We observed a significant delay in RMCT among patients with C8/T1 radiculopathy when compared to control subjects, and also demonstrated that this delay was independently associated with the presence of radiculopathy at the C8/T1 level. This finding is notable because it provided a rational basis for the applicability of RMCT in terms of diagnosing radiculopathy at a specific spinal root level.\\nThe application of RMCT to assess radiculopathy occurring at particular levels is rare. Most relevant studies were conducted before the 2000s, with only a few published since then. Banerjee et al. [13] described magnetic spinal stimulation as a non-invasive method to evaluate lumbosacral motor radiculopathy in 26 patients and 25 control subjects; they mainly targeted the lower lumbosacral nerve roots (L5, S1, and S2) by recording from the abductor hallucis muscle. According to their results, patients with clinical motor weakness had profoundly prolonged RMCTs; therefore, they suggested that RMCT delay was related to the severity of symptoms. Golez [22] measured motor conduction time via stimulation at the L1 and S1 levels in 25 patients with lumbar spinal stenosis and 36 control subjects. That study was noteworthy because it demonstrated that neurogenic claudication increased the cauda equina motor conduction time, revealing that the clinical symptom and delay of motor conduction time were related. In our study, the RMCT showed a significant correlation with the duration of subjective symptoms, which supports Golez’s results to some extent. However, in our study, the subjective pain level in the lesional-side limb was not related to RMCT as we assumed when we primarily evaluated the motor nerve. Additionally, Seçil et al. [23] measured the cauda equina motor conduction time and showed that it was slower in the lumbar spinal stenosis group than in the control group; however, their study did not identify the relationship between the degree of motor conduction delay and symptom severity.\\nOur study is the first to measure RMCT in patients with cervical radiculopathy and has the advantage of being specific to a specific spinal level; moreover, we included a relatively large number of subjects compared to previous related studies. Additionally, variables such as age, sex, and height were analyzed and adjusted for using PSM, rendering the results much more reliable. Since the APB-RMCT was previously found to be affected by height in a linear model [18], for the integrity of our results, it was important that we controlled for this factor when assessing the patient and control groups.\\nRMCT has also been studied in patients with demyelinating neuropathy. In a previous study, a significant difference was found in the RMCT of 30 healthy subjects and 12 patients with Guillain-Barre syndrome who were diagnosed within one week [14]; hence, RMCT was deemed to be useful for the early diagnosis of focal segmental demyelinating polyneuropathy, affording this measure a novel clinical utility. In the study of Inaba et al. [24], RMCT was measured in 11 patients with chronic inflammatory demyelinating neuropathy and 10 with Guillain-Barre syndrome, wherein the authors found that the RMCT increasingly normalized as muscle strength recovered. Taken together, such studies illustrate the applicability of the RMCT to various diseases involving the motor root segment. Furthermore, measuring the RMCT is a non-invasive procedure and has the potential to be applied to additional fields given its ability to detect diseases in their early stages and to also assess functional recovery. Therefore, additional research on the use of the RMCT is warranted going forward.\\nThis study had several limitations. It was retrospective in nature, requiring additional investigations to validate our results. Additionally, our participants were limited to Koreans; hence, validation among patients of other ethnicities is also required. Our patients comprised a single group diagnosed with C8/T1 radiculopathy that was relatively heterogeneous; future studies that analyze various subgroups of patients (such as those with acute versus chronic lesions or mild versus severe symptoms), would further clarify the relationship between RMCT and radiculopathy. Finally, there were potential limitations in terms of measurement. For example, it is difficult to apply our methods at some spinal root levels because reliable recording of the F-wave is possible only from distal muscles [25]. Moreover, it was challenging to measure the RMCT when reliable motor evoked potentials or compound muscle action potentials could not be obtained owing to severe peripheral nerve lesions in the distal limbs.\",\n \"We demonstrated that the RMCT is delayed in the root lesions of patients with C8/T1 level radiculopathy. This noninvasive method (compared to EMG) can have an adjuvant role in diagnosing radiculopathy at certain spinal levels. Our study may be a good milestone for future clinical applications of RMCT as related to patients with radiculopathy. However, it is important to validate our results through additional multi-center, multi-ethnic studies.\",\n \"[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\nBelow is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction",null,null,null,null,"results",null,null,"discussion","conclusion","supplementary-material",null],"string":"[\n \"introduction\",\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n \"discussion\",\n \"conclusion\",\n \"supplementary-material\",\n null\n]"},"keywords":{"kind":"list like","value":["Radiculopathy","Electrodiagnosis","Motor-evoked potential","Nerve conduction"],"string":"[\n \"Radiculopathy\",\n \"Electrodiagnosis\",\n \"Motor-evoked potential\",\n \"Nerve conduction\"\n]"}}},{"rowIdx":372,"cells":{"title":{"kind":"string","value":"Overwork among resident physicians: national questionnaire survey results."},"pmid":{"kind":"string","value":"36266644"},"background_abstract":{"kind":"string","value":"Residents experience the longest working hours among physicians. Thus, it would be beneficial to perform a nationwide survey in Japan on residents' long work hours and the background factors promoting upper limits on working hours of Japanese residents. The aim of this study was to study or assess the state of physicians' excessive work hours and its background factors using a questionnaire survey."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"The survey was sent to 924 hospitals. The physicians' general attributes, work hours and conditions, and employers' foundational entities were explored. Multiple logistic regression analysis was performed to elucidate the background factors for long work hours."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Of the 4306 resident physicians who responded, 67% had ≥ 60 in-hospital hours/week and 27% had ≥ 80 h/week; 51% were on-call ≥ four times/month. Many of them hoped for increased remuneration. Additionally, female (reference: male, OR: 0.65, 95% CI: 0.55-0.76), 35-40 years old (reference: 25-30 years old, OR: 1.83, 95% CI: 1.32-2.54), childlessness (reference: child, OR: 1.41, 95% CI: 1.12-1.75), surgical specialization (reference: internal medicine, OR: 2.51, 95% CI: 1.96-3.23), neurosurgical specialization (reference: internal medicine, OR: 4.38, 95% CI: 2.92-6.59) and hospitals with 200-400 physicians (reference: <100 physicians, OR: 1.82, 95% CI: 1.12-2.96) exhibited significant correlations with ≥ 80 in-hospital hours/week."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Understanding the factors that increase the likelihood of residents working very long hours could aid in making targeted changes to address the specific concerns. Moreover, reducing working hours to a reasonable limit can improve resident physicians' health and the quality of care they provide in their community."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Adult","Female","Humans","Male","Internal Medicine","Internship and Residency","Japan","Physicians","Surveys and Questionnaires","Workload"],"string":"[\n \"Adult\",\n \"Female\",\n \"Humans\",\n \"Male\",\n \"Internal Medicine\",\n \"Internship and Residency\",\n \"Japan\",\n \"Physicians\",\n \"Surveys and Questionnaires\",\n \"Workload\"\n]"},"pmcid":{"kind":"string","value":"9584270"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"Regarding the respondents’ attributes shown in Tables 1 and 37% were female, 60% were between 30 − 39 years of age, 44% were married, and 21% had children. Additionally, 23% specialized in general internal medicine. The average annual income ranged from ≥ 4 to < 6 million yen. Regarding employer attributes, 54% and 38% worked in large urban areas and national university hospitals, respectively. The number of beds was between 600 and 800, and there were ≥ 600 physicians in most hospitals. Moreover, 70% of the physicians had a part-time job.\n\nTable 1Respondent attributes and in-hospital hoursAttributes of Respondentsn%Annual Income (not including part-time work; dollars)n (%)Sex< 16,0001383.2Male271062.9≥ 16,000–<32,000101423.5Female159637.1≥ 32,000–<48,000108325.2Age (years)≥ 48,000–<64,00087920.4≥ 25–<30258760.1≥ 64,000–<80,00067315.6≥ 30–<35127029.5≥ 80,000–<96,0003438.0≥ 35–<402205.1≥ 96,0001764.1≥ 402295.3In-hospital (hours per week)Marital Status< 40 h781.8Married190444.240–60 h130530.3Not married240255.860–80 h178341.4Parental Status80–100 h78318.2Children90921.1≥ 100 h3578.3No children339778.9Foundational Entity of EmployerBasic Area of SpecializationPublic122428.4General Internal Medicine97922.7National University161837.6Surgery43710.1Private University94521.9Orthopedic Surgery3297.6Private51912.1Pediatrics2976.9Employer’s Total No. of BedsOB/GYN2746.4< 200631.5Anesthesiology2726.3≥ 200–<4003207.4Otolaryngology2275.3≥ 400–<60067715.7Psychiatry2124.9≥ 600–<800141232.8Dermatology1874.3≥ 800–<1,000107224.9Ophthalmology1523.5≥ 1,00076217.7Emergency Medicine1463.4Employer’s No. of Full-time PhysiciansUrology1453.4< 1004219.8Radiology1393.2≥ 100–<20076817.8Neurosurgery1252.9≥ 200–<30059213.7Pathology1182.7≥ 300–<40073317.0Plastic Surgery1052.4≥ 400–<50047611.1General Practice892.1≥ 6001,31630.6Rehabilitation Medicine591.4Employer’s Regional ClassificationClinical Testing140.3Urban231553.8Part-time WorkIntermediate191444.4Yes299169.5Rural771.8No131530.5\n\nRespondent attributes and in-hospital hours\nAs for the weekly in-hospital time, 32%, 41%, and 27% of physicians spent < 60, 60–80 h, and greater than 80 h in the hospital per week, respectively.\nTable 2 shows the results of multiple logistic regression analysis with in-hospital time of ≥ 80 h per week as the explanatory variable in order to elucidate the background factors for long work hours of residents.\n\nTable 2Multiple logistic regression analysis with residents’ long work hours as the explanatory variableOdds Ratio95% Confidence IntervalP ValueSexMaleReferenceFemale0.650.55–0.76< 0.01Age (years)≥ 25–<30Reference≥ 30–<351.241.05–1.460.01≥ 35–<401.831.32–2.54< 0.01≥ 400.770.51–1.150.20Marital StatusMarriedReferenceNot married1.110.93–1.320.23Parental StatusChildrenReferenceNo children1.411.12–1.75< 0.01Basic Area of SpecializationGeneral Internal MedicineReferenceSurgery2.511.96–3.23< 0.01Orthopedic Surgery1.290.97–1.740.08Pediatrics1.320.98–1.780.07OB/GYN1.360.99–1.870.06Anesthesiology0.460.31–0.69< 0.01Otolaryngology0.690.47–1.000.05Psychiatry0.570.38–0.85< 0.01Dermatology0.690.45–1.050.08Ophthalmology0.610.39–0.960.03Emergency Medicine1.110.74–1.660.62Urology1.370.93–2.030.11Radiology0.280.15–0.51< 0.01Neurosurgery4.382.92–6.59< 0.01Pathology0.460.26–0.81< 0.01Plastic Surgery1.220.77–1.940.40General Practice0.810.45–1.470.49Rehabilitation Medicine0.500.23–1.080.08Clinical Testing0.240.03–1.950.18Annual Income (dollars)< 16,000Reference≥ 16,000–<32,0000.840.54–1.300.43≥ 32,000–<48,0000.970.62–1.520.90≥ 48,000–<64,0000.790.50–1.250.31≥ 64,000–<80,0000.960.60–1.530.86≥ 80,000–<96,0000.710.42–1.170.18≥ 96,0000.760.42–1.360.36Foundational Entity of EmployerPublicReferenceNational University0.850.62–1.160.30Private University0.960.66–1.400.84Private1.000.76–1.320.99Employer’s Total No. of Beds< 200Reference≥ 200–<4001.450.65–3.210.36≥ 400–<6001.330.58–3.040.50≥ 600–<8001.400.59–3.310.45≥ 800–<1,0001.370.57–3.320.48≥ 1,0001.430.59–3.510.43Employer’s No. of Full-time Physicians< 100Reference≥ 100–<2001.250.86–1.820.25≥ 200–<3001.641.05–2.550.03≥ 300–<4001.821.12–2.960.02≥ 400–<5001.560.93–2.620.10≥ 6001.410.85–2.340.18Employer’s Regional ClassificationUrbanReferenceIntermediate1.000.85–1.180.98Rural0.820.41–1.620.56\n\nMultiple logistic regression analysis with residents’ long work hours as the explanatory variable\nSex, age (≥ 30–<40 years), no children, surgical/neurosurgical specialization, and number of physicians (≥ 200–<400) exhibited statistically significant correlation as background factors for physicians being in-hospital for ≥ 80 h per week. However, there were no statistically significant correlations with marital status, income, foundational entity of employer, number of beds, and regional characteristics."},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"Our national survey shows that many residents work very long hours (more than 80 h/week), with some specialists working even longer hours, while others are required to work part-time jobs in order to be financially stable. Moreover, our results show a variety of background factors that correlate to working very long hours, such as being male, being between 30 and 40 years old, having no children, being in the surgical or neurosurgical specialty, and the number of physicians in the hospital. By understanding the factors that increase the likelihood of residents working very long hours, targeted changes can be made to address the specific concerns. Reducing working hours to a reasonable level can improve the resident physicians’ health and the quality of care they provide in their community."},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Survey participants","Investigation items","Statistical analysis","Ethical considerations"],"string":"[\n \"Background\",\n \"Methods\",\n \"Survey participants\",\n \"Investigation items\",\n \"Statistical analysis\",\n \"Ethical considerations\"\n]"},"other_sections_texts":{"kind":"list like","value":["The work hours of Japanese workers are indicated to be longer compared to workers in other countries [1]. The long working hours of physicians, approximately more than 60 h a week, account for 42% of full-time physicians working more than 200 days per year, compared to a 14% average for all other professions [2]. Among physicians, residents (physicians undergoing training in order to become specialists) have the longest working hours [3].\nDepression-related suicides and deaths due to ischemic heart disease and cerebrovascular disease caused by overwork are called “karoshi” (death due to overwork) and have become a public health issue specific to East Asia, such as China [4]. Moreover, there is ample evidence regarding the adverse effects of long work hours on health, both in Japan and overseas. According to a systematic review by Bannai et al. [5], long working hours are correlated with depression, anxiety, sleep, and coronary artery disease.\nIn 2003, the US Accreditation Council for Graduate Medical Education (ACGME) limited residents’ work hours to less than 80 h per week due to its adverse effects on their health and the increased risk of medical error [6, 7]. A systematic review that evaluated this measure concluded that the work-hour restrictions of the ACGME are correlated with an improvement in emotional malaise and burnout syndrome symptoms [8].\nMoreover, in Japan, several studies performed at single facilities have indicated the depressive tendencies and burnout of interns and residents [9, 10]. However, unlike those in the US, there have been no restrictions on working hours. Additionally, some physicians not being paid [11] and long working hours [3] have been indicated as issues in the working environment of Japanese residents. However, no nationwide survey studies have been conducted in Japan on resident physicians.\nIn 2016, the Ministry of Health, Labour and Welfare instituted the “Study Meeting on Physician Labor Reform,” which has been examining measures to promote labor reforms for physicians, and it has disclosed a policy wherein from April 2024, the upper limit on annual overtime work hours for residents will be 1860 h (at par with the US ACGME) [12].\nTo administratively promote the setting of upper limits on working hours for Japanese residents, it will be useful to perform a nationwide survey on the condition of long work hours for residents and its background factors, and examine topics relevant to the promotion of physician labor reforms.\nTherefore, a nationwide questionnaire survey was performed on Japanese residents who work particularly long hours, and it discussed the overworked condition of the resident physicians and background factors associated with its effects to suggest policies for the promotion of labor reforms.","[SUBTITLE] Survey participants [SUBSECTION] Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\nAccounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\n[SUBTITLE] Investigation items [SUBSECTION] A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\nA web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\n[SUBTITLE] Statistical analysis [SUBSECTION] To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\nTo elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\n[SUBTITLE] Ethical considerations [SUBSECTION] This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\nThis study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.","Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.","A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.","To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.","This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information."],"string":"[\n \"The work hours of Japanese workers are indicated to be longer compared to workers in other countries [1]. The long working hours of physicians, approximately more than 60 h a week, account for 42% of full-time physicians working more than 200 days per year, compared to a 14% average for all other professions [2]. Among physicians, residents (physicians undergoing training in order to become specialists) have the longest working hours [3].\\nDepression-related suicides and deaths due to ischemic heart disease and cerebrovascular disease caused by overwork are called “karoshi” (death due to overwork) and have become a public health issue specific to East Asia, such as China [4]. Moreover, there is ample evidence regarding the adverse effects of long work hours on health, both in Japan and overseas. According to a systematic review by Bannai et al. [5], long working hours are correlated with depression, anxiety, sleep, and coronary artery disease.\\nIn 2003, the US Accreditation Council for Graduate Medical Education (ACGME) limited residents’ work hours to less than 80 h per week due to its adverse effects on their health and the increased risk of medical error [6, 7]. A systematic review that evaluated this measure concluded that the work-hour restrictions of the ACGME are correlated with an improvement in emotional malaise and burnout syndrome symptoms [8].\\nMoreover, in Japan, several studies performed at single facilities have indicated the depressive tendencies and burnout of interns and residents [9, 10]. However, unlike those in the US, there have been no restrictions on working hours. Additionally, some physicians not being paid [11] and long working hours [3] have been indicated as issues in the working environment of Japanese residents. However, no nationwide survey studies have been conducted in Japan on resident physicians.\\nIn 2016, the Ministry of Health, Labour and Welfare instituted the “Study Meeting on Physician Labor Reform,” which has been examining measures to promote labor reforms for physicians, and it has disclosed a policy wherein from April 2024, the upper limit on annual overtime work hours for residents will be 1860 h (at par with the US ACGME) [12].\\nTo administratively promote the setting of upper limits on working hours for Japanese residents, it will be useful to perform a nationwide survey on the condition of long work hours for residents and its background factors, and examine topics relevant to the promotion of physician labor reforms.\\nTherefore, a nationwide questionnaire survey was performed on Japanese residents who work particularly long hours, and it discussed the overworked condition of the resident physicians and background factors associated with its effects to suggest policies for the promotion of labor reforms.\",\n \"[SUBTITLE] Survey participants [SUBSECTION] Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\\nAccounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\\n[SUBTITLE] Investigation items [SUBSECTION] A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\\nA web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\\n[SUBTITLE] Statistical analysis [SUBSECTION] To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\\nTo elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\\n[SUBTITLE] Ethical considerations [SUBSECTION] This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\\nThis study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\",\n \"Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\",\n \"A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\",\n \"To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\",\n \"This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Survey participants","Investigation items","Statistical analysis","Ethical considerations","Results","Discussion","Conclusion"],"string":"[\n \"Background\",\n \"Methods\",\n \"Survey participants\",\n \"Investigation items\",\n \"Statistical analysis\",\n \"Ethical considerations\",\n \"Results\",\n \"Discussion\",\n \"Conclusion\"\n]"},"all_sections_texts":{"kind":"list like","value":["The work hours of Japanese workers are indicated to be longer compared to workers in other countries [1]. The long working hours of physicians, approximately more than 60 h a week, account for 42% of full-time physicians working more than 200 days per year, compared to a 14% average for all other professions [2]. Among physicians, residents (physicians undergoing training in order to become specialists) have the longest working hours [3].\nDepression-related suicides and deaths due to ischemic heart disease and cerebrovascular disease caused by overwork are called “karoshi” (death due to overwork) and have become a public health issue specific to East Asia, such as China [4]. Moreover, there is ample evidence regarding the adverse effects of long work hours on health, both in Japan and overseas. According to a systematic review by Bannai et al. [5], long working hours are correlated with depression, anxiety, sleep, and coronary artery disease.\nIn 2003, the US Accreditation Council for Graduate Medical Education (ACGME) limited residents’ work hours to less than 80 h per week due to its adverse effects on their health and the increased risk of medical error [6, 7]. A systematic review that evaluated this measure concluded that the work-hour restrictions of the ACGME are correlated with an improvement in emotional malaise and burnout syndrome symptoms [8].\nMoreover, in Japan, several studies performed at single facilities have indicated the depressive tendencies and burnout of interns and residents [9, 10]. However, unlike those in the US, there have been no restrictions on working hours. Additionally, some physicians not being paid [11] and long working hours [3] have been indicated as issues in the working environment of Japanese residents. However, no nationwide survey studies have been conducted in Japan on resident physicians.\nIn 2016, the Ministry of Health, Labour and Welfare instituted the “Study Meeting on Physician Labor Reform,” which has been examining measures to promote labor reforms for physicians, and it has disclosed a policy wherein from April 2024, the upper limit on annual overtime work hours for residents will be 1860 h (at par with the US ACGME) [12].\nTo administratively promote the setting of upper limits on working hours for Japanese residents, it will be useful to perform a nationwide survey on the condition of long work hours for residents and its background factors, and examine topics relevant to the promotion of physician labor reforms.\nTherefore, a nationwide questionnaire survey was performed on Japanese residents who work particularly long hours, and it discussed the overworked condition of the resident physicians and background factors associated with its effects to suggest policies for the promotion of labor reforms.","[SUBTITLE] Survey participants [SUBSECTION] Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\nAccounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\n[SUBTITLE] Investigation items [SUBSECTION] A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\nA web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\n[SUBTITLE] Statistical analysis [SUBSECTION] To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\nTo elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\n[SUBTITLE] Ethical considerations [SUBSECTION] This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\nThis study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.","Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.","A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.","To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.","This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.","Regarding the respondents’ attributes shown in Tables 1 and 37% were female, 60% were between 30 − 39 years of age, 44% were married, and 21% had children. Additionally, 23% specialized in general internal medicine. The average annual income ranged from ≥ 4 to < 6 million yen. Regarding employer attributes, 54% and 38% worked in large urban areas and national university hospitals, respectively. The number of beds was between 600 and 800, and there were ≥ 600 physicians in most hospitals. Moreover, 70% of the physicians had a part-time job.\n\nTable 1Respondent attributes and in-hospital hoursAttributes of Respondentsn%Annual Income (not including part-time work; dollars)n (%)Sex< 16,0001383.2Male271062.9≥ 16,000–<32,000101423.5Female159637.1≥ 32,000–<48,000108325.2Age (years)≥ 48,000–<64,00087920.4≥ 25–<30258760.1≥ 64,000–<80,00067315.6≥ 30–<35127029.5≥ 80,000–<96,0003438.0≥ 35–<402205.1≥ 96,0001764.1≥ 402295.3In-hospital (hours per week)Marital Status< 40 h781.8Married190444.240–60 h130530.3Not married240255.860–80 h178341.4Parental Status80–100 h78318.2Children90921.1≥ 100 h3578.3No children339778.9Foundational Entity of EmployerBasic Area of SpecializationPublic122428.4General Internal Medicine97922.7National University161837.6Surgery43710.1Private University94521.9Orthopedic Surgery3297.6Private51912.1Pediatrics2976.9Employer’s Total No. of BedsOB/GYN2746.4< 200631.5Anesthesiology2726.3≥ 200–<4003207.4Otolaryngology2275.3≥ 400–<60067715.7Psychiatry2124.9≥ 600–<800141232.8Dermatology1874.3≥ 800–<1,000107224.9Ophthalmology1523.5≥ 1,00076217.7Emergency Medicine1463.4Employer’s No. of Full-time PhysiciansUrology1453.4< 1004219.8Radiology1393.2≥ 100–<20076817.8Neurosurgery1252.9≥ 200–<30059213.7Pathology1182.7≥ 300–<40073317.0Plastic Surgery1052.4≥ 400–<50047611.1General Practice892.1≥ 6001,31630.6Rehabilitation Medicine591.4Employer’s Regional ClassificationClinical Testing140.3Urban231553.8Part-time WorkIntermediate191444.4Yes299169.5Rural771.8No131530.5\n\nRespondent attributes and in-hospital hours\nAs for the weekly in-hospital time, 32%, 41%, and 27% of physicians spent < 60, 60–80 h, and greater than 80 h in the hospital per week, respectively.\nTable 2 shows the results of multiple logistic regression analysis with in-hospital time of ≥ 80 h per week as the explanatory variable in order to elucidate the background factors for long work hours of residents.\n\nTable 2Multiple logistic regression analysis with residents’ long work hours as the explanatory variableOdds Ratio95% Confidence IntervalP ValueSexMaleReferenceFemale0.650.55–0.76< 0.01Age (years)≥ 25–<30Reference≥ 30–<351.241.05–1.460.01≥ 35–<401.831.32–2.54< 0.01≥ 400.770.51–1.150.20Marital StatusMarriedReferenceNot married1.110.93–1.320.23Parental StatusChildrenReferenceNo children1.411.12–1.75< 0.01Basic Area of SpecializationGeneral Internal MedicineReferenceSurgery2.511.96–3.23< 0.01Orthopedic Surgery1.290.97–1.740.08Pediatrics1.320.98–1.780.07OB/GYN1.360.99–1.870.06Anesthesiology0.460.31–0.69< 0.01Otolaryngology0.690.47–1.000.05Psychiatry0.570.38–0.85< 0.01Dermatology0.690.45–1.050.08Ophthalmology0.610.39–0.960.03Emergency Medicine1.110.74–1.660.62Urology1.370.93–2.030.11Radiology0.280.15–0.51< 0.01Neurosurgery4.382.92–6.59< 0.01Pathology0.460.26–0.81< 0.01Plastic Surgery1.220.77–1.940.40General Practice0.810.45–1.470.49Rehabilitation Medicine0.500.23–1.080.08Clinical Testing0.240.03–1.950.18Annual Income (dollars)< 16,000Reference≥ 16,000–<32,0000.840.54–1.300.43≥ 32,000–<48,0000.970.62–1.520.90≥ 48,000–<64,0000.790.50–1.250.31≥ 64,000–<80,0000.960.60–1.530.86≥ 80,000–<96,0000.710.42–1.170.18≥ 96,0000.760.42–1.360.36Foundational Entity of EmployerPublicReferenceNational University0.850.62–1.160.30Private University0.960.66–1.400.84Private1.000.76–1.320.99Employer’s Total No. of Beds< 200Reference≥ 200–<4001.450.65–3.210.36≥ 400–<6001.330.58–3.040.50≥ 600–<8001.400.59–3.310.45≥ 800–<1,0001.370.57–3.320.48≥ 1,0001.430.59–3.510.43Employer’s No. of Full-time Physicians< 100Reference≥ 100–<2001.250.86–1.820.25≥ 200–<3001.641.05–2.550.03≥ 300–<4001.821.12–2.960.02≥ 400–<5001.560.93–2.620.10≥ 6001.410.85–2.340.18Employer’s Regional ClassificationUrbanReferenceIntermediate1.000.85–1.180.98Rural0.820.41–1.620.56\n\nMultiple logistic regression analysis with residents’ long work hours as the explanatory variable\nSex, age (≥ 30–<40 years), no children, surgical/neurosurgical specialization, and number of physicians (≥ 200–<400) exhibited statistically significant correlation as background factors for physicians being in-hospital for ≥ 80 h per week. However, there were no statistically significant correlations with marital status, income, foundational entity of employer, number of beds, and regional characteristics.","First, 68% and 27% of the responding physicians exceeded 60 and 80 h of in-hospital time per week, respectively, elucidating the actual state of overwork for Japanese resident physicians. An in-hospital time of 60 h per week converts to four hours of overtime per day for an 8-hours/day 40-hour week. This is the legal work hour stipulated by the Labor Standards Act and generates over 80 h of monthly overtime work.\nLong work hours that continue for several months are called the “death by overwork level” because of the strong correlation with mental disorders and the onset of circulatory organ diseases, and it is a criterion for the recognition of labor accidents [14, 15]. This study’s results suggest that 68% of the respondents work in an environment exceeding the “death by overwork level.”\nThe Japanese Ministry of Health, Labour and Welfare decided to implement “Physician Labor Reform” in April 2024; the principle annual upper limit of overtime work for physicians shall be 960 h, 1860 h in specific cases, and various initiatives to achieve this goal are underway [16]. The plan promotes shifting tasks to other hospital staff and sets a continuous work time limit (28 h) with 9-hour intervals between work shifts and an upper limit for overtime hours [12].\nFor the 1860-hour cases, the average weekly in-hospital time was estimated to be < 80 h. This study’s results suggested that 27% of respondents worked hours that exceeded this criterion. Because such long working hours will become illegal in April 2024, measures must be taken to shorten working hours as soon as possible. However, the background factors for such super-long work hours have not been identified.\nLong work hours of ≥ 80 in-hospital hours per week exhibited a significant correlation with males and no children, whereas no correlation with marital status was observed. The presence of children may require shorter work hours due to the time spent on childcare. Another possibility is that long work hours are not inhibited by the time invested in married life. Regarding sex, a report on American female surgery residents suggests that they have a higher proportion of long work hours and higher rates of experiencing burnout [16]. However, in this study, males had a stronger correlation with > 80 h of work per week. This is affected by the higher proportion of Japanese males in specialized areas that have long work hours, such as surgery and neurosurgery [17]. In this study, when controlling for specialty on multivariate analyses, male gender was still associated with long work hours.\nRegarding age, the odds ratio was higher at ≥ 35–<40 years of age; 5% of participants belonged to this age group. These individuals may have become physicians later, or become residents again after changing their specialization, but the correlation with excessively long work hours is unclear. Further studies, such as additional surveys and interviews, will be necessary to explain why these physicians often work excessively long hours and identify other background factors.\nDifferences in long work hours were based on specialization areas. For example, surgeons often have long work hours. The Japan Surgical Society has stated that promoting initiatives such as drastic task shifting and consolidation of surgeries are necessary to shorten overtime hours. However, deeper discussions on self-improvement are necessary to ensure that reducing work hours does not impair surgeons’ acquisition of knowledge and skills [18]. In several countries, discussions are occurring regarding restricting work hours and whether policies that promote uniform work hour restrictions, regardless of specialization, are not appropriate [19].\nIn Japan, each physician can select their specialization both in medical school and after graduation [20]. Therefore, it is difficult to eliminate the uneven distribution of physician specialties. Physicians may also transfer from their selected, busy specialty to a less busy one [21]. According to the Japanese Ministry of Health, Labour and Welfare, the lowest proportion of physicians is in surgery and neurosurgery, where this study’s results showed significantly longer working hours. In contrast, there is a surplus in ophthalmology and dermatology, where long work hours are significantly scarce.\nNo significant correlation was observed between ≥ 80 in-hospital hours per week and the residents’ income. As stated previously, the annual pay of residents is low, and some physicians are even unpaid [11]. Therefore, because physicians’ long work hours are not necessarily compensated by greater pay, annual income may not be regarded as a background factor.\nThe regulations for residents’ work hours in the “Labor Reforms for Physicians” being advanced by the Japanese Ministry of Health, Labour and Welfare are based on those implemented by the US ACGME [22]. These regulations comprise 80 working hours per week, a continuous working time limit of 16 h, and a 10-hour interval between working hours. Contrastingly, the Japanese reforms comprise an upper limit of 1860 h on annual overtime, a 28-hour limit on continuous working time, and a 9-hour interval between working hours.\nIn the US, various evidence-based reviews have been performed on the ACGME regulations on residents’ work hours. Comparative reports regarding residents in surgery and internal medicine, using an intervention group without the continuous work time and shift interval restrictions and a control group using the ACGME regulated restrictions, indicated no significant difference in patient outcomes or residents’ degrees of satisfaction [23, 24]. The two studies cited were large. However, they had significant limitations, such as power to detect changes in mortality and outcomes metrics indirectly linked to study populations of interest, and the findings generally ran counter to the broader literature on the effects of efforts to reduce work hours on patient safety.\nAnother study in 2005 shows extended-duration work shifts (over 24 h), which were sanctioned by ACGME, pose safety hazards for interns [25]. In 2011, ACGME instituted a continuous working time limit of 16 h for residents. A recent study shows the 2011 ACGME work-hour limit was associated with meaningful improvements in physician safety and health [26]. Implementing such studies and examining evidence-based systemic reviews from the perspectives of patient safety, physician health, physician training, and so forth, is necessary in Japan.\nJapanese resident physicians tend to have a low annual income. In this study, 52% of respondents had an annual income of less than 48,000 dollars (not including part-time work), and 70% were working part-time. Part-time workers are assumed to need to engage in part-time work at medical institutions other than their regular employer to be financially stable or to repay loans. Additionally, some doctors work at university hospitals as “unpaid doctors” [11]. Therefore, the need to improve residents’ remuneration can also be regarded as an issue.\nIf part-time work is treated as one of the causes of residents’ overwork, then the reduction of part-time work would be one method of decreasing residents’ work hours by promoting labor reforms for physicians. However, this may further reduce residents’ incomes. Therefore, there should also be proposals for pay supplementation using the national budget.\nAdditionally, medical institutions that have maintained their treatment systems so far through the part-time employment of residents must remember that treatment systems may become unsustainable with the termination of part-time work. Moreover, consideration should be given at the policy level so that local medical care is not affected, particularly in regions with a shortage of doctors.\nThis study has several limitations. First is the possibility of selection bias due to the individuals’ voluntary participation. Despite this limitation, effective responses were obtained from 4306 participants at 416 hospitals out of the 924 core hospitals nationwide accredited by the Japanese Medical Specialty Board (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. Moreover, both the response rate and the number of respondents ensured a sample that was representative of the population.\nSecond, the survey used self-report responses, which may have introduced various information biases. For example, misreporting was possible because the in-hospital hours per week were not a detailed time study filled out by another person. Therefore, it couldn’t be confirmed whether the in-hospital hours were all work hours. Additionally, because participation in this survey was voluntary, detailed definitions of the specialist jargon used were not provided in order to increase the response rate.\nThird, a statistical correlation was observed between long work hours and background factors. However, the causal relationship was unclear. It is possible that probable confounding factors were not measured in this study. Additionally, the effects of the COVID-19 pandemic were not examined. As shown in previous studies, the COVID-19 pandemic has greatly affected how physicians work [27], and it would be useful to conduct the same survey after the COVID-19 pandemic and compare the results of the studies. Furthermore, because there have been no previous similar nationwide studies on Japanese residents, it is difficult to compare the results of this study with the situation of residents before the COVID-19 pandemic.\nThe harsh working conditions of resident physicians in Japan and the background factors of longer work hours have been elucidated in tandem with discussions of solutions. To ensure the health of resident physicians, the quality of training, and the safety of patients, it is necessary to ascertain the status of resident physicians in greater detail, including vigorously promoting and verifying the effects of labor reforms for physicians.","Our national survey shows that many residents work very long hours (more than 80 h/week), with some specialists working even longer hours, while others are required to work part-time jobs in order to be financially stable. Moreover, our results show a variety of background factors that correlate to working very long hours, such as being male, being between 30 and 40 years old, having no children, being in the surgical or neurosurgical specialty, and the number of physicians in the hospital. By understanding the factors that increase the likelihood of residents working very long hours, targeted changes can be made to address the specific concerns. Reducing working hours to a reasonable level can improve the resident physicians’ health and the quality of care they provide in their community."],"string":"[\n \"The work hours of Japanese workers are indicated to be longer compared to workers in other countries [1]. The long working hours of physicians, approximately more than 60 h a week, account for 42% of full-time physicians working more than 200 days per year, compared to a 14% average for all other professions [2]. Among physicians, residents (physicians undergoing training in order to become specialists) have the longest working hours [3].\\nDepression-related suicides and deaths due to ischemic heart disease and cerebrovascular disease caused by overwork are called “karoshi” (death due to overwork) and have become a public health issue specific to East Asia, such as China [4]. Moreover, there is ample evidence regarding the adverse effects of long work hours on health, both in Japan and overseas. According to a systematic review by Bannai et al. [5], long working hours are correlated with depression, anxiety, sleep, and coronary artery disease.\\nIn 2003, the US Accreditation Council for Graduate Medical Education (ACGME) limited residents’ work hours to less than 80 h per week due to its adverse effects on their health and the increased risk of medical error [6, 7]. A systematic review that evaluated this measure concluded that the work-hour restrictions of the ACGME are correlated with an improvement in emotional malaise and burnout syndrome symptoms [8].\\nMoreover, in Japan, several studies performed at single facilities have indicated the depressive tendencies and burnout of interns and residents [9, 10]. However, unlike those in the US, there have been no restrictions on working hours. Additionally, some physicians not being paid [11] and long working hours [3] have been indicated as issues in the working environment of Japanese residents. However, no nationwide survey studies have been conducted in Japan on resident physicians.\\nIn 2016, the Ministry of Health, Labour and Welfare instituted the “Study Meeting on Physician Labor Reform,” which has been examining measures to promote labor reforms for physicians, and it has disclosed a policy wherein from April 2024, the upper limit on annual overtime work hours for residents will be 1860 h (at par with the US ACGME) [12].\\nTo administratively promote the setting of upper limits on working hours for Japanese residents, it will be useful to perform a nationwide survey on the condition of long work hours for residents and its background factors, and examine topics relevant to the promotion of physician labor reforms.\\nTherefore, a nationwide questionnaire survey was performed on Japanese residents who work particularly long hours, and it discussed the overworked condition of the resident physicians and background factors associated with its effects to suggest policies for the promotion of labor reforms.\",\n \"[SUBTITLE] Survey participants [SUBSECTION] Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\\nAccounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\\n[SUBTITLE] Investigation items [SUBSECTION] A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\\nA web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\\n[SUBTITLE] Statistical analysis [SUBSECTION] To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\\nTo elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\\n[SUBTITLE] Ethical considerations [SUBSECTION] This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\\nThis study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\",\n \"Accounting for the 19 basic areas stipulated by the Japanese Medical Specialty Board, the survey was sent to 924 core hospitals nationwide that had received training program approval for each of the basic areas. Effective responses were received from 4306 residents at 416 hospitals (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. The survey was conducted for 14 days, from 10 October to 23 October 2020.\",\n \"A web survey was conducted using a questionnaire. First, the respondents’ attributes (sex, age, basic area of specialization, marital/parental status, income (1 dollar = 125 yen), and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications) were described. There were four types of foundational entities of the employers: public (except for national universities); national universities; private universities; and private (except for private universities). The regional classifications were divided into Group 1 (urban areas), Group 2 (intermediate areas), and Group 3 (rural areas) based on the combination of the population size and density as of 2019 regarding the 344 secondary care zones nationwide [13].\\nNext, with regard to the employment status of the residents, the number of hours spent in the hospital, the number of times on duty, etc. were described.\",\n \"To elucidate the background factors for residents’ long work hours, multiple logistic regression analyses was performed with ≥ 80 in-hospital hours per week as the explanatory variable and sex, age, basic field of specialization, marital/parental status, income, and foundational entity of employer/no. of beds/no. of full-time doctors/regional classifications as response variables. In the statistical analysis, a p value of < 0.05 was considered statistically significant. STATA 15.1 was used in all analyses.\",\n \"This study was conducted after receiving approval from the medical ethics board of the University of Tsukuba Faculty of Medicine (no. 1498). All methods were performed in accordance with the relevant guidelines and regulations of the Declaration of Helsinki. On the first page of the questionnaire, the purpose of the study and measures for safe data management were explained. Additionally, participants in this study were informed that their participation was entirely voluntary. Data were collected from participants who provided informed consent to participate. The results of the survey were analyzed separately in order to ensure the anonymization and confidentiality of personal information.\",\n \"Regarding the respondents’ attributes shown in Tables 1 and 37% were female, 60% were between 30 − 39 years of age, 44% were married, and 21% had children. Additionally, 23% specialized in general internal medicine. The average annual income ranged from ≥ 4 to < 6 million yen. Regarding employer attributes, 54% and 38% worked in large urban areas and national university hospitals, respectively. The number of beds was between 600 and 800, and there were ≥ 600 physicians in most hospitals. Moreover, 70% of the physicians had a part-time job.\\n\\nTable 1Respondent attributes and in-hospital hoursAttributes of Respondentsn%Annual Income (not including part-time work; dollars)n (%)Sex< 16,0001383.2Male271062.9≥ 16,000–<32,000101423.5Female159637.1≥ 32,000–<48,000108325.2Age (years)≥ 48,000–<64,00087920.4≥ 25–<30258760.1≥ 64,000–<80,00067315.6≥ 30–<35127029.5≥ 80,000–<96,0003438.0≥ 35–<402205.1≥ 96,0001764.1≥ 402295.3In-hospital (hours per week)Marital Status< 40 h781.8Married190444.240–60 h130530.3Not married240255.860–80 h178341.4Parental Status80–100 h78318.2Children90921.1≥ 100 h3578.3No children339778.9Foundational Entity of EmployerBasic Area of SpecializationPublic122428.4General Internal Medicine97922.7National University161837.6Surgery43710.1Private University94521.9Orthopedic Surgery3297.6Private51912.1Pediatrics2976.9Employer’s Total No. of BedsOB/GYN2746.4< 200631.5Anesthesiology2726.3≥ 200–<4003207.4Otolaryngology2275.3≥ 400–<60067715.7Psychiatry2124.9≥ 600–<800141232.8Dermatology1874.3≥ 800–<1,000107224.9Ophthalmology1523.5≥ 1,00076217.7Emergency Medicine1463.4Employer’s No. of Full-time PhysiciansUrology1453.4< 1004219.8Radiology1393.2≥ 100–<20076817.8Neurosurgery1252.9≥ 200–<30059213.7Pathology1182.7≥ 300–<40073317.0Plastic Surgery1052.4≥ 400–<50047611.1General Practice892.1≥ 6001,31630.6Rehabilitation Medicine591.4Employer’s Regional ClassificationClinical Testing140.3Urban231553.8Part-time WorkIntermediate191444.4Yes299169.5Rural771.8No131530.5\\n\\nRespondent attributes and in-hospital hours\\nAs for the weekly in-hospital time, 32%, 41%, and 27% of physicians spent < 60, 60–80 h, and greater than 80 h in the hospital per week, respectively.\\nTable 2 shows the results of multiple logistic regression analysis with in-hospital time of ≥ 80 h per week as the explanatory variable in order to elucidate the background factors for long work hours of residents.\\n\\nTable 2Multiple logistic regression analysis with residents’ long work hours as the explanatory variableOdds Ratio95% Confidence IntervalP ValueSexMaleReferenceFemale0.650.55–0.76< 0.01Age (years)≥ 25–<30Reference≥ 30–<351.241.05–1.460.01≥ 35–<401.831.32–2.54< 0.01≥ 400.770.51–1.150.20Marital StatusMarriedReferenceNot married1.110.93–1.320.23Parental StatusChildrenReferenceNo children1.411.12–1.75< 0.01Basic Area of SpecializationGeneral Internal MedicineReferenceSurgery2.511.96–3.23< 0.01Orthopedic Surgery1.290.97–1.740.08Pediatrics1.320.98–1.780.07OB/GYN1.360.99–1.870.06Anesthesiology0.460.31–0.69< 0.01Otolaryngology0.690.47–1.000.05Psychiatry0.570.38–0.85< 0.01Dermatology0.690.45–1.050.08Ophthalmology0.610.39–0.960.03Emergency Medicine1.110.74–1.660.62Urology1.370.93–2.030.11Radiology0.280.15–0.51< 0.01Neurosurgery4.382.92–6.59< 0.01Pathology0.460.26–0.81< 0.01Plastic Surgery1.220.77–1.940.40General Practice0.810.45–1.470.49Rehabilitation Medicine0.500.23–1.080.08Clinical Testing0.240.03–1.950.18Annual Income (dollars)< 16,000Reference≥ 16,000–<32,0000.840.54–1.300.43≥ 32,000–<48,0000.970.62–1.520.90≥ 48,000–<64,0000.790.50–1.250.31≥ 64,000–<80,0000.960.60–1.530.86≥ 80,000–<96,0000.710.42–1.170.18≥ 96,0000.760.42–1.360.36Foundational Entity of EmployerPublicReferenceNational University0.850.62–1.160.30Private University0.960.66–1.400.84Private1.000.76–1.320.99Employer’s Total No. of Beds< 200Reference≥ 200–<4001.450.65–3.210.36≥ 400–<6001.330.58–3.040.50≥ 600–<8001.400.59–3.310.45≥ 800–<1,0001.370.57–3.320.48≥ 1,0001.430.59–3.510.43Employer’s No. of Full-time Physicians< 100Reference≥ 100–<2001.250.86–1.820.25≥ 200–<3001.641.05–2.550.03≥ 300–<4001.821.12–2.960.02≥ 400–<5001.560.93–2.620.10≥ 6001.410.85–2.340.18Employer’s Regional ClassificationUrbanReferenceIntermediate1.000.85–1.180.98Rural0.820.41–1.620.56\\n\\nMultiple logistic regression analysis with residents’ long work hours as the explanatory variable\\nSex, age (≥ 30–<40 years), no children, surgical/neurosurgical specialization, and number of physicians (≥ 200–<400) exhibited statistically significant correlation as background factors for physicians being in-hospital for ≥ 80 h per week. However, there were no statistically significant correlations with marital status, income, foundational entity of employer, number of beds, and regional characteristics.\",\n \"First, 68% and 27% of the responding physicians exceeded 60 and 80 h of in-hospital time per week, respectively, elucidating the actual state of overwork for Japanese resident physicians. An in-hospital time of 60 h per week converts to four hours of overtime per day for an 8-hours/day 40-hour week. This is the legal work hour stipulated by the Labor Standards Act and generates over 80 h of monthly overtime work.\\nLong work hours that continue for several months are called the “death by overwork level” because of the strong correlation with mental disorders and the onset of circulatory organ diseases, and it is a criterion for the recognition of labor accidents [14, 15]. This study’s results suggest that 68% of the respondents work in an environment exceeding the “death by overwork level.”\\nThe Japanese Ministry of Health, Labour and Welfare decided to implement “Physician Labor Reform” in April 2024; the principle annual upper limit of overtime work for physicians shall be 960 h, 1860 h in specific cases, and various initiatives to achieve this goal are underway [16]. The plan promotes shifting tasks to other hospital staff and sets a continuous work time limit (28 h) with 9-hour intervals between work shifts and an upper limit for overtime hours [12].\\nFor the 1860-hour cases, the average weekly in-hospital time was estimated to be < 80 h. This study’s results suggested that 27% of respondents worked hours that exceeded this criterion. Because such long working hours will become illegal in April 2024, measures must be taken to shorten working hours as soon as possible. However, the background factors for such super-long work hours have not been identified.\\nLong work hours of ≥ 80 in-hospital hours per week exhibited a significant correlation with males and no children, whereas no correlation with marital status was observed. The presence of children may require shorter work hours due to the time spent on childcare. Another possibility is that long work hours are not inhibited by the time invested in married life. Regarding sex, a report on American female surgery residents suggests that they have a higher proportion of long work hours and higher rates of experiencing burnout [16]. However, in this study, males had a stronger correlation with > 80 h of work per week. This is affected by the higher proportion of Japanese males in specialized areas that have long work hours, such as surgery and neurosurgery [17]. In this study, when controlling for specialty on multivariate analyses, male gender was still associated with long work hours.\\nRegarding age, the odds ratio was higher at ≥ 35–<40 years of age; 5% of participants belonged to this age group. These individuals may have become physicians later, or become residents again after changing their specialization, but the correlation with excessively long work hours is unclear. Further studies, such as additional surveys and interviews, will be necessary to explain why these physicians often work excessively long hours and identify other background factors.\\nDifferences in long work hours were based on specialization areas. For example, surgeons often have long work hours. The Japan Surgical Society has stated that promoting initiatives such as drastic task shifting and consolidation of surgeries are necessary to shorten overtime hours. However, deeper discussions on self-improvement are necessary to ensure that reducing work hours does not impair surgeons’ acquisition of knowledge and skills [18]. In several countries, discussions are occurring regarding restricting work hours and whether policies that promote uniform work hour restrictions, regardless of specialization, are not appropriate [19].\\nIn Japan, each physician can select their specialization both in medical school and after graduation [20]. Therefore, it is difficult to eliminate the uneven distribution of physician specialties. Physicians may also transfer from their selected, busy specialty to a less busy one [21]. According to the Japanese Ministry of Health, Labour and Welfare, the lowest proportion of physicians is in surgery and neurosurgery, where this study’s results showed significantly longer working hours. In contrast, there is a surplus in ophthalmology and dermatology, where long work hours are significantly scarce.\\nNo significant correlation was observed between ≥ 80 in-hospital hours per week and the residents’ income. As stated previously, the annual pay of residents is low, and some physicians are even unpaid [11]. Therefore, because physicians’ long work hours are not necessarily compensated by greater pay, annual income may not be regarded as a background factor.\\nThe regulations for residents’ work hours in the “Labor Reforms for Physicians” being advanced by the Japanese Ministry of Health, Labour and Welfare are based on those implemented by the US ACGME [22]. These regulations comprise 80 working hours per week, a continuous working time limit of 16 h, and a 10-hour interval between working hours. Contrastingly, the Japanese reforms comprise an upper limit of 1860 h on annual overtime, a 28-hour limit on continuous working time, and a 9-hour interval between working hours.\\nIn the US, various evidence-based reviews have been performed on the ACGME regulations on residents’ work hours. Comparative reports regarding residents in surgery and internal medicine, using an intervention group without the continuous work time and shift interval restrictions and a control group using the ACGME regulated restrictions, indicated no significant difference in patient outcomes or residents’ degrees of satisfaction [23, 24]. The two studies cited were large. However, they had significant limitations, such as power to detect changes in mortality and outcomes metrics indirectly linked to study populations of interest, and the findings generally ran counter to the broader literature on the effects of efforts to reduce work hours on patient safety.\\nAnother study in 2005 shows extended-duration work shifts (over 24 h), which were sanctioned by ACGME, pose safety hazards for interns [25]. In 2011, ACGME instituted a continuous working time limit of 16 h for residents. A recent study shows the 2011 ACGME work-hour limit was associated with meaningful improvements in physician safety and health [26]. Implementing such studies and examining evidence-based systemic reviews from the perspectives of patient safety, physician health, physician training, and so forth, is necessary in Japan.\\nJapanese resident physicians tend to have a low annual income. In this study, 52% of respondents had an annual income of less than 48,000 dollars (not including part-time work), and 70% were working part-time. Part-time workers are assumed to need to engage in part-time work at medical institutions other than their regular employer to be financially stable or to repay loans. Additionally, some doctors work at university hospitals as “unpaid doctors” [11]. Therefore, the need to improve residents’ remuneration can also be regarded as an issue.\\nIf part-time work is treated as one of the causes of residents’ overwork, then the reduction of part-time work would be one method of decreasing residents’ work hours by promoting labor reforms for physicians. However, this may further reduce residents’ incomes. Therefore, there should also be proposals for pay supplementation using the national budget.\\nAdditionally, medical institutions that have maintained their treatment systems so far through the part-time employment of residents must remember that treatment systems may become unsustainable with the termination of part-time work. Moreover, consideration should be given at the policy level so that local medical care is not affected, particularly in regions with a shortage of doctors.\\nThis study has several limitations. First is the possibility of selection bias due to the individuals’ voluntary participation. Despite this limitation, effective responses were obtained from 4306 participants at 416 hospitals out of the 924 core hospitals nationwide accredited by the Japanese Medical Specialty Board (response rate = 49%). The response rate is related to the % of hospitals from which the responses were received. The number of residents and the location and specialty are unknown. Moreover, both the response rate and the number of respondents ensured a sample that was representative of the population.\\nSecond, the survey used self-report responses, which may have introduced various information biases. For example, misreporting was possible because the in-hospital hours per week were not a detailed time study filled out by another person. Therefore, it couldn’t be confirmed whether the in-hospital hours were all work hours. Additionally, because participation in this survey was voluntary, detailed definitions of the specialist jargon used were not provided in order to increase the response rate.\\nThird, a statistical correlation was observed between long work hours and background factors. However, the causal relationship was unclear. It is possible that probable confounding factors were not measured in this study. Additionally, the effects of the COVID-19 pandemic were not examined. As shown in previous studies, the COVID-19 pandemic has greatly affected how physicians work [27], and it would be useful to conduct the same survey after the COVID-19 pandemic and compare the results of the studies. Furthermore, because there have been no previous similar nationwide studies on Japanese residents, it is difficult to compare the results of this study with the situation of residents before the COVID-19 pandemic.\\nThe harsh working conditions of resident physicians in Japan and the background factors of longer work hours have been elucidated in tandem with discussions of solutions. To ensure the health of resident physicians, the quality of training, and the safety of patients, it is necessary to ascertain the status of resident physicians in greater detail, including vigorously promoting and verifying the effects of labor reforms for physicians.\",\n \"Our national survey shows that many residents work very long hours (more than 80 h/week), with some specialists working even longer hours, while others are required to work part-time jobs in order to be financially stable. Moreover, our results show a variety of background factors that correlate to working very long hours, such as being male, being between 30 and 40 years old, having no children, being in the surgical or neurosurgical specialty, and the number of physicians in the hospital. By understanding the factors that increase the likelihood of residents working very long hours, targeted changes can be made to address the specific concerns. Reducing working hours to a reasonable level can improve the resident physicians’ health and the quality of care they provide in their community.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,"results","discussion","conclusion"],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n \"discussion\",\n \"conclusion\"\n]"},"keywords":{"kind":"list like","value":["Physicians","Burnout","Questionnaire survey","Japan","Work hours"],"string":"[\n \"Physicians\",\n \"Burnout\",\n \"Questionnaire survey\",\n \"Japan\",\n \"Work hours\"\n]"}}},{"rowIdx":373,"cells":{"title":{"kind":"string","value":"The value of combining the simple anthropometric obesity parameters, Body Mass Index (BMI) and a Body Shape Index (ABSI), to assess the risk of non-alcoholic fatty liver disease."},"pmid":{"kind":"string","value":"36266655"},"background_abstract":{"kind":"string","value":"Body mass index (BMI) and A Body Shape Index (ABSI) are current independent risk factors for non-alcoholic fatty liver disease (NAFLD). The aim of this study was to explore the value of combining these two most common obesity indexes in identifying NAFLD."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"The subjects in this study were 14,251 individuals from the NAfld in the Gifu Area, Longitudinal Analysis (NAGALA) cohort who underwent routine health examination. We integrated BMI with WC and with ABSI to construct 6 combined obesity indicators-obesity phenotypes, the combined anthropometric risk index (ARI) for BMI and ABSI, optimal proportional combination OBMI+WC and OBMI+ABSI, and multiplicative combination BMI*WC and BMI*ABSI. Several multivariable logistic regression models were established to evaluate the relationship between BMI, WC, ABSI, and the above six combined indicators and NAFLD; receiver operating characteristic (ROC) curves were drawn to compare the ability of each obesity indicator to identify NAFLD."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"A total of 2,507 (17.59%) subjects were diagnosed with NAFLD. BMI, WC, ABSI, and all other combined obesity indicators were significantly and positively associated with NAFLD in the current study, with BMI*WC having the strongest correlation with NAFLD in female subjects (OR per SD increase: 3.13) and BMI*ABSI having the strongest correlation in male subjects (OR per SD increase: 2.97). ROC analysis showed that ARI and OBMI+ABSI had the best diagnostic performance in both sexes, followed by BMI*WC (area under the curve: female 0.8912; male 0.8270). After further age stratification, it was found that ARI and multiplicative indicators (BMI*WC, BMI*ABSI) and optimal proportional combination indicators (OBMI+WC, OBMI+ABSI) significantly improved the NAFLD risk identification ability of the basic anthropometric parameters in middle-aged females and young and middle-aged males."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"In the general population, BMI combined with ABSI best identified obesity-related NAFLD risk and was significantly better than BMI or WC, or ABSI. We find that ARI and the multiplicative combined indicators BMI*WC and BMI*ABSI further improved risk prediction and may be proposed for possible use in clinical practice."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Male","Female","Body Mass Index","Non-alcoholic Fatty Liver Disease","Waist Circumference","Anthropometry","Obesity","Obesity, Morbid","Risk Factors"],"string":"[\n \"Humans\",\n \"Male\",\n \"Female\",\n \"Body Mass Index\",\n \"Non-alcoholic Fatty Liver Disease\",\n \"Waist Circumference\",\n \"Anthropometry\",\n \"Obesity\",\n \"Obesity, Morbid\",\n \"Risk Factors\"\n]"},"pmcid":{"kind":"string","value":"9585710"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Demographic and clinical characteristics of subjects [SUBSECTION] The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1\nThe study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1\n[SUBTITLE] Calculation of the optimal proportional combination of BMI with WC and ABSI [SUBSECTION] In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1\nIn the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1\n[SUBTITLE] Association of various obesity indicators with NAFLD [SUBSECTION] Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\nCorrelation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n[SUBTITLE] Accuracy of various obesity indicators to identify NAFLD [SUBSECTION] ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders\nROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"Taken together, our findings confirmed that BMI combined with ABSI and WC can better explain obesity-related NAFLD risk than a single indicator in the general population, and identified the ARI and the multiplicative combination indicators BMI*ABSI and BMI*WC as the simplest and very effective combination indicators for diagnosing the risk of NAFLD. These new findings underscored the importance of the combined application of BMI and the abdominal obesity index in clinical practice to assess NAFLD risk in the general population and provided a cost-effective tool for precise preventive screening and treatment monitoring for NAFLD."},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Data sources and study design","Definition and derivation of the combined indicators of BMI with WC and ABSI","Combination using usual cut-offs for BMI and WC","Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI","Diagnosis of NAFLD","Statistical analysis","Demographic and clinical characteristics of subjects","Calculation of the optimal proportional combination of BMI with WC and ABSI","Association of various obesity indicators with NAFLD","Accuracy of various obesity indicators to identify NAFLD","Advantages and limitations of research",""],"string":"[\n \"Background\",\n \"Methods\",\n \"Data sources and study design\",\n \"Definition and derivation of the combined indicators of BMI with WC and ABSI\",\n \"Combination using usual cut-offs for BMI and WC\",\n \"Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI\",\n \"Diagnosis of NAFLD\",\n \"Statistical analysis\",\n \"Demographic and clinical characteristics of subjects\",\n \"Calculation of the optimal proportional combination of BMI with WC and ABSI\",\n \"Association of various obesity indicators with NAFLD\",\n \"Accuracy of various obesity indicators to identify NAFLD\",\n \"Advantages and limitations of research\",\n \"\"\n]"},"other_sections_texts":{"kind":"list like","value":["Non-alcoholic fatty liver disease (NAFLD) is a chronic non-infectious liver disease characterized by oxidative stress, inflammation, and fibrosis in hepatocytes among people without a history of excessive alcohol consumption [1, 2]. NAFLD can present as asymptomatic simple hepatic steatosis in its early stages, and without intervention, oxidative stress, inflammation, and fibrosis in hepatocytes can contribute to the continuous progression of NAFLD to non-alcoholic steatohepatitis and then to cirrhosis and ultimately to hepatocellular carcinoma [2, 3]. Recent epidemiological surveys have shown that about a quarter of the global population suffers from NAFLD, with a 27.4% estimated prevalence in Asia [4]. NAFLD is not only the most common chronic liver disease worldwide but has replaced hepatitis B virus infection as the primary risk factor for advanced liver diseases such as cirrhosis and hepatocellular carcinoma [5]. The high prevalence of NAFLD and its severe complications place a heavy burden on humans and the world’s health care systems, making the early prevention, identification, and intervention of NAFLD of great public health importance.\nObesity is one of the most important risk factors for NAFLD [6]. Waist circumference (WC) and body mass index (BMI) are currently the most widely used basic body measures for the assessment of general and central obesity and also are known risk factors for NAFLD [7–10]. Both BMI and WC alone, however, have some distinct limitations, such as the inability of BMI alone to distinguish between adipose and muscle tissue and the inability of WC to determine the differential contribution of abdominal visceral adipose tissue and abdominal subcutaneous adipose tissue to central obesity [11–13]. It is not clear if combining risk associated with BMI and WC can improve the individual risk association. Several relevant studies have analyzed the performance of the combination of BMI and WC in disease risk assessment, and overall, combining BMI and WC in different ways significantly improved the predictive efficacy for obesity-related diseases such as obesity-related hypertension, type 2 diabetes, cardiovascular disease, and all-cause mortality risk compared to a single indicator [14–17]. In addition, a recent study by Wang et al. also found that high levels of WC and BMI were both associated with NAFLD risk in people with type 2 diabetes and that BMI combined with WC was superior to the single measures in assessing the risk of NAFLD [18]. What remains unclear, however, is whether the combination of BMI and WC best identifies a higher risk of NAFLD in the general population. Furthermore, it is important to note that there is a strong correlation between BMI and WC (r ≈ 0.80), and directly combining BMI with WC has the potential to confound the association between BMI and NAFLD, leading to biased and possibly misleading risk estimates [19–21]. A Body Shape Index (ABSI), an allometric abdominal obesity index calculated from WC, height, and weight, is intended to be independent of BMI and has proven to be useful for identifying individuals at risk for sarcopenic obesity [22]; therefore, we also evaluated combining BMI with the non-correlated abdominal obesity index ABSI. This study sought to explore risk assessment for NAFLD of various combinations of BMI and WC or ABSI in a cohort of general health examinees from NAGALA.","[SUBTITLE] Data sources and study design [SUBSECTION] In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\nIn the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\n[SUBTITLE] Definition and derivation of the combined indicators of BMI with WC and ABSI [SUBSECTION] Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\nBased on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\n[SUBTITLE] Combination using usual cut-offs for BMI and WC [SUBSECTION] In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\nIn accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\n[SUBTITLE] Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI [SUBSECTION] The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\nThe optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\n[SUBTITLE] Diagnosis of NAFLD [SUBSECTION] Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\nSpecialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\n[SUBTITLE] Statistical analysis [SUBSECTION] We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\nWe performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.","In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).","Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].","In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.","The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.","Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].","We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.","The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1","In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1","Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.","ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders","Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\n\nSupplementary Material 2\n\nSupplementary Material 2"],"string":"[\n \"Non-alcoholic fatty liver disease (NAFLD) is a chronic non-infectious liver disease characterized by oxidative stress, inflammation, and fibrosis in hepatocytes among people without a history of excessive alcohol consumption [1, 2]. NAFLD can present as asymptomatic simple hepatic steatosis in its early stages, and without intervention, oxidative stress, inflammation, and fibrosis in hepatocytes can contribute to the continuous progression of NAFLD to non-alcoholic steatohepatitis and then to cirrhosis and ultimately to hepatocellular carcinoma [2, 3]. Recent epidemiological surveys have shown that about a quarter of the global population suffers from NAFLD, with a 27.4% estimated prevalence in Asia [4]. NAFLD is not only the most common chronic liver disease worldwide but has replaced hepatitis B virus infection as the primary risk factor for advanced liver diseases such as cirrhosis and hepatocellular carcinoma [5]. The high prevalence of NAFLD and its severe complications place a heavy burden on humans and the world’s health care systems, making the early prevention, identification, and intervention of NAFLD of great public health importance.\\nObesity is one of the most important risk factors for NAFLD [6]. Waist circumference (WC) and body mass index (BMI) are currently the most widely used basic body measures for the assessment of general and central obesity and also are known risk factors for NAFLD [7–10]. Both BMI and WC alone, however, have some distinct limitations, such as the inability of BMI alone to distinguish between adipose and muscle tissue and the inability of WC to determine the differential contribution of abdominal visceral adipose tissue and abdominal subcutaneous adipose tissue to central obesity [11–13]. It is not clear if combining risk associated with BMI and WC can improve the individual risk association. Several relevant studies have analyzed the performance of the combination of BMI and WC in disease risk assessment, and overall, combining BMI and WC in different ways significantly improved the predictive efficacy for obesity-related diseases such as obesity-related hypertension, type 2 diabetes, cardiovascular disease, and all-cause mortality risk compared to a single indicator [14–17]. In addition, a recent study by Wang et al. also found that high levels of WC and BMI were both associated with NAFLD risk in people with type 2 diabetes and that BMI combined with WC was superior to the single measures in assessing the risk of NAFLD [18]. What remains unclear, however, is whether the combination of BMI and WC best identifies a higher risk of NAFLD in the general population. Furthermore, it is important to note that there is a strong correlation between BMI and WC (r ≈ 0.80), and directly combining BMI with WC has the potential to confound the association between BMI and NAFLD, leading to biased and possibly misleading risk estimates [19–21]. A Body Shape Index (ABSI), an allometric abdominal obesity index calculated from WC, height, and weight, is intended to be independent of BMI and has proven to be useful for identifying individuals at risk for sarcopenic obesity [22]; therefore, we also evaluated combining BMI with the non-correlated abdominal obesity index ABSI. This study sought to explore risk assessment for NAFLD of various combinations of BMI and WC or ABSI in a cohort of general health examinees from NAGALA.\",\n \"[SUBTITLE] Data sources and study design [SUBSECTION] In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\\nIn the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\\n[SUBTITLE] Definition and derivation of the combined indicators of BMI with WC and ABSI [SUBSECTION] Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\\nBased on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\\n[SUBTITLE] Combination using usual cut-offs for BMI and WC [SUBSECTION] In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\\nIn accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\\n[SUBTITLE] Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI [SUBSECTION] The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\\nThe optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\\n[SUBTITLE] Diagnosis of NAFLD [SUBSECTION] Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\\nSpecialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\\n[SUBTITLE] Statistical analysis [SUBSECTION] We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\\nWe performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\",\n \"In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\",\n \"Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\",\n \"In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\",\n \"The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\",\n \"Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\",\n \"We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\",\n \"The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\\n\\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\n\\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\\n\\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\\n\\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\\nAbbreviations as in Table 1\",\n \"In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\\n\\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\\n\\nOptimal combination equations of BMI with WC and ABSI.\\nAbbreviations as in Table 1\",\n \"Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\\n\\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\n\\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\\nModel 1 adjusted for age and height\\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\",\n \"ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\\n\\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\\n\\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\\nABSI**\\nARI**\\n0.6171\\n0.8912\\n0.5916\\n0.8775\\n0.6426\\n0.9048\\n75.7925\\n12.7825\\n0.5926\\n0.7735\\n0.5962\\n0.8577\\n0.1888\\n0.6312\\nAbbreviations as in Table 1\\n**P < 0.05 compared with BMI*WC in each gender\\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\\n\\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\\n\\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\\nAbbreviations as in Table 1\\n***P < 0.05 compared with BMI*WC in each age group of both genders\",\n \"Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\n\\nSupplementary Material 2\\n\\nSupplementary Material 2\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Data sources and study design","Definition and derivation of the combined indicators of BMI with WC and ABSI","Combination using usual cut-offs for BMI and WC","Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI","Diagnosis of NAFLD","Statistical analysis","Results","Demographic and clinical characteristics of subjects","Calculation of the optimal proportional combination of BMI with WC and ABSI","Association of various obesity indicators with NAFLD","Accuracy of various obesity indicators to identify NAFLD","Discussion","Advantages and limitations of research","Conclusion","Electronic supplementary material",""],"string":"[\n \"Background\",\n \"Methods\",\n \"Data sources and study design\",\n \"Definition and derivation of the combined indicators of BMI with WC and ABSI\",\n \"Combination using usual cut-offs for BMI and WC\",\n \"Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI\",\n \"Diagnosis of NAFLD\",\n \"Statistical analysis\",\n \"Results\",\n \"Demographic and clinical characteristics of subjects\",\n \"Calculation of the optimal proportional combination of BMI with WC and ABSI\",\n \"Association of various obesity indicators with NAFLD\",\n \"Accuracy of various obesity indicators to identify NAFLD\",\n \"Discussion\",\n \"Advantages and limitations of research\",\n \"Conclusion\",\n \"Electronic supplementary material\",\n \"\"\n]"},"all_sections_texts":{"kind":"list like","value":["Non-alcoholic fatty liver disease (NAFLD) is a chronic non-infectious liver disease characterized by oxidative stress, inflammation, and fibrosis in hepatocytes among people without a history of excessive alcohol consumption [1, 2]. NAFLD can present as asymptomatic simple hepatic steatosis in its early stages, and without intervention, oxidative stress, inflammation, and fibrosis in hepatocytes can contribute to the continuous progression of NAFLD to non-alcoholic steatohepatitis and then to cirrhosis and ultimately to hepatocellular carcinoma [2, 3]. Recent epidemiological surveys have shown that about a quarter of the global population suffers from NAFLD, with a 27.4% estimated prevalence in Asia [4]. NAFLD is not only the most common chronic liver disease worldwide but has replaced hepatitis B virus infection as the primary risk factor for advanced liver diseases such as cirrhosis and hepatocellular carcinoma [5]. The high prevalence of NAFLD and its severe complications place a heavy burden on humans and the world’s health care systems, making the early prevention, identification, and intervention of NAFLD of great public health importance.\nObesity is one of the most important risk factors for NAFLD [6]. Waist circumference (WC) and body mass index (BMI) are currently the most widely used basic body measures for the assessment of general and central obesity and also are known risk factors for NAFLD [7–10]. Both BMI and WC alone, however, have some distinct limitations, such as the inability of BMI alone to distinguish between adipose and muscle tissue and the inability of WC to determine the differential contribution of abdominal visceral adipose tissue and abdominal subcutaneous adipose tissue to central obesity [11–13]. It is not clear if combining risk associated with BMI and WC can improve the individual risk association. Several relevant studies have analyzed the performance of the combination of BMI and WC in disease risk assessment, and overall, combining BMI and WC in different ways significantly improved the predictive efficacy for obesity-related diseases such as obesity-related hypertension, type 2 diabetes, cardiovascular disease, and all-cause mortality risk compared to a single indicator [14–17]. In addition, a recent study by Wang et al. also found that high levels of WC and BMI were both associated with NAFLD risk in people with type 2 diabetes and that BMI combined with WC was superior to the single measures in assessing the risk of NAFLD [18]. What remains unclear, however, is whether the combination of BMI and WC best identifies a higher risk of NAFLD in the general population. Furthermore, it is important to note that there is a strong correlation between BMI and WC (r ≈ 0.80), and directly combining BMI with WC has the potential to confound the association between BMI and NAFLD, leading to biased and possibly misleading risk estimates [19–21]. A Body Shape Index (ABSI), an allometric abdominal obesity index calculated from WC, height, and weight, is intended to be independent of BMI and has proven to be useful for identifying individuals at risk for sarcopenic obesity [22]; therefore, we also evaluated combining BMI with the non-correlated abdominal obesity index ABSI. This study sought to explore risk assessment for NAFLD of various combinations of BMI and WC or ABSI in a cohort of general health examinees from NAGALA.","[SUBTITLE] Data sources and study design [SUBSECTION] In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\nIn the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\n[SUBTITLE] Definition and derivation of the combined indicators of BMI with WC and ABSI [SUBSECTION] Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\nBased on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\n[SUBTITLE] Combination using usual cut-offs for BMI and WC [SUBSECTION] In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\nIn accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\n[SUBTITLE] Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI [SUBSECTION] The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\nThe optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\n[SUBTITLE] Diagnosis of NAFLD [SUBSECTION] Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\nSpecialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\n[SUBTITLE] Statistical analysis [SUBSECTION] We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\nWe performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.","In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\n\nFig. 1Flowchart of the selection process of study subjects\n\nFlowchart of the selection process of study subjects\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).","Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].","In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.","The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.","Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].","We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.","[SUBTITLE] Demographic and clinical characteristics of subjects [SUBSECTION] The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1\nThe study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1\n[SUBTITLE] Calculation of the optimal proportional combination of BMI with WC and ABSI [SUBSECTION] In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1\nIn the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1\n[SUBTITLE] Association of various obesity indicators with NAFLD [SUBSECTION] Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\nCorrelation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n[SUBTITLE] Accuracy of various obesity indicators to identify NAFLD [SUBSECTION] ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders\nROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders","The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\n\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\n\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\n\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\n\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\nAbbreviations as in Table 1","In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\n\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\n\nOptimal combination equations of BMI with WC and ABSI.\nAbbreviations as in Table 1","Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\n\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\n\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\nModel 1 adjusted for age and height\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.","ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\n\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\n\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\n\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\nABSI**\nARI**\n0.6171\n0.8912\n0.5916\n0.8775\n0.6426\n0.9048\n75.7925\n12.7825\n0.5926\n0.7735\n0.5962\n0.8577\n0.1888\n0.6312\nAbbreviations as in Table 1\n**P < 0.05 compared with BMI*WC in each gender\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\n\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\n\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\nAbbreviations as in Table 1\n***P < 0.05 compared with BMI*WC in each age group of both genders","Our main findings showed that BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, and the obesity phenotypes were all significantly and positively correlated with NAFLD risk; and for the first time in the general population, it was found that BMI combined with WC and ABSI can significantly improve the ability of the basic anthropometric measures to identify NAFLD, especially in middle-aged females and young and middle-aged males. Of significant mention, compared with other age stages in both sexes, all obesity indicators in this study have the highest diagnostic efficiency among young females.\nIn recent years, the rapid global economic development, dietary patterns and lifestyles of the residents have changed dramatically, and a high-calorie, high-fat daily diet and a sedentary lifestyle have become the main themes of today’s society; therefore, the prevalence of metabolic syndrome, obesity, and other obesity-related diseases has increased greatly [2, 32, 33]. NAFLD is considered the hepatic manifestation of the metabolic syndrome and encompasses a complex spectrum of liver pathologies ranging from simple hepatic steatosis to non-alcoholic steatohepatitis, and also the most common chronic liver disease and the leading cause of liver-related death worldwide [1, 34]. The occurrence and development of NAFLD are closely related to the global epidemic of obesity [6]. Recent epidemiological surveys have shown that the incidence of NAFLD in obese patients can reach 80%, and in severely obese patients who require bariatric surgery, the incidence of NAFLD is as high as 90% [4].\nIndependent associations between NAFLD and traditional obesity indicators BMI and WC have been demonstrated in many observational studies, but there is much debate as to which indicator is better at identifying obesity-related NAFLD risk [7–10, 35–39]. Several recent studies have demonstrated that BMI was a stronger indicator of obesity for identifying and predicting NAFLD risk, with BMI having the largest AUC compared with other obesity-related parameters, showing better diagnostic performance than WC and other obesity indicators [37–39]. Of course, there were other studies that expressed the opposite standpoints; two studies of NAFLD in obese adolescents and children found that both WC and BMI independently predicted NAFLD risk, with WC, in particular, having the highest predictive accuracy [35, 36]. Furthermore, a meta-analysis study by Qing et al. found that WC remained the strongest anthropometric predictor of NAFLD even after adjusting for a large number of confounding factors, including BMI, and that central obesity was a more dangerous pattern of body fat distribution than general obesity [7]. Clearly, both BMI and WC were independent risk factors for NAFLD, but neither BMI nor WC alone appeared to be an adequate measure of obesity-related NAFLD risk. In a study by Janssen et al. [40], it was found that either BMI or WC alone was an independent predictor of total fat, non-abdominal fat, abdominal subcutaneous fat, and visceral fat, but 20–40% of the variation in fat mass in these fat pools remained unexplained by BMI or WC alone; after they further combining BMI with WC, found that the variability of fat content in these fat pools decreased significantly. Given that the available evidence generally supports both abdominal subcutaneous fat and visceral fat as independent predictors of insulin resistance [41, 42], we hypothesized that the combination of WC and BMI may be a better predictor of metabolic disease. This speculation has been confirmed in studies of metabolic diseases such as cardiovascular disease, obesity-related hypertension, type 2 diabetes, and risk of all-cause mortality [14–17]. Furthermore, in a recent study, Wang et al. also found that BMI combined with WC was more strongly associated with NAFLD in a diabetic population than either BMI or WC alone [18]. However, their study subjects were limited to the diabetic population and did not explore the value of combined BMI and WC in the identification of NAFLD. In addition, it is also worth noting that in the above-mentioned studies related to BMI combined with WC, almost all researchers ignored the statistical artifacts that may result from the strong correlation between BMI and WC. A recent study by Christakoudi S et al. found a traditional U-shaped association between BMI alone and all-cause mortality risk, but when BMI was combined with WC, which had a strong correlation, a predominantly negative association was found between BMI and all-cause mortality, i.e., the higher the BMI the lower the mortality risk [19]. It means that the combination of WC and BMI may not simply supplement BMI, but even changed the pattern of associations between BMI and outcome events, which may be a statistical artifact caused by the strong correlation between WC and BMI. In addition, when they combined BMI with the moderately correlated waist index waist-to-height ratio found a similar but more modest change in the pattern of association between BMI and all-cause mortality. Therefore, combining BMI with a correlated abdominal obesity index may lead to biased and possibly even misleading risk estimates and ineffective risk stratification [19]. As expected, when the researchers further combined BMI with the unrelated abdominal obesity index ABSI and hip circumference, they found that the pattern of association between BMI and all-cause mortality did not change and that more fine-grained risk stratification was achieved for ABSI in each BMI stratum. ABSI is a reliable predictor of diseases such as cardiovascular disease and all-cause mortality and a valid proxy for identifying sarcopenic obesity, independent of BMI by design, and is a useful complement to BMI for risk estimation and risk stratification [22, 43–45].\nIn this study, we also analyzed the correlations between BMI and WC and ABSI, and in agreement with the results of previous studies [19], there was a strong correlation between BMI and WC and almost no correlation between BMI and ABSI (r < 0.05); in addition, in the association analysis of single indicators with NAFLD risk we found that BMI, WC, and ABSI were all significantly and positively associated with the risk of NAFLD (Table 4). Therefore, the combination of BMI and ABSI may be more appropriate than the combination of BMI and WC. Nevertheless, to fully explore the value of combining BMI and the abdominal obesity index to assess the risk of NAFLD and to circumvent possible statistical artifacts, we still combined BMI with WC and ABSI in 4 separate ways based on previous studies: First, the combination of BMI and WC into four obesity phenotypes, BMIN/WCN, BMIO/WCN, BMIN/WCO, and BMIO/WCO, based on the overweight cut-offs for WC and BMI in Asian populations recommended by the WHO Expert Committee [27, 28]. Second, constructing optimal proportional combinations OBMI+WC and OBMI+ABSI based on the regression coefficients of BMI, WC, and ABSI in multivariate logistic regression models. Third, multiplied BMI with WC and ABSI directly to obtain BMI*WC and BMI*ABSI. Fourth, a predictor ARI (BMI, ABSI) was calculated using the method developed by Krakauer NY et al. [26]. In the analysis of the association between each obesity phenotype and NAFLD we found that after adjusting for a large number of confounders (Model 3), subjects with normal BMI and WC in both sexes had the lowest risk of NAFLD, both the BMIO/WCN phenotype and the BMIN/WCO phenotype increased the risk of NAFLD, and subjects with the BMIO/WCO phenotype had the highest risk of NAFLD (male: OR 4.19, 95%CI 3.54, 4.96; female: OR 9.78, 95%CI 7.35, 13.02); these results were in accord with previous studies [18]. Obviously, comprehensive consideration of the effects of BMI and WC on NAFLD in the general population could explain the obesity-related NAFLD risk to a greater extent, identifying more populations at risk for developing NAFLD. Additionally, all other combined indicators in this study were significantly and positively correlated with NAFLD, with BMI*ABSI having the strongest correlation among males and BMI*WC having the strongest correlation with NAFLD among females. The results also did not change after treating BMI*ABSI as a categorical variable, but it is worth mentioning that the Q4 category has a much higher risk of NAFLD in females than in males compared with the Q1 category in BMI*ABSI (OR: female 63.97 vs. male 26.15). Consistently, in the obesity phenotypes, female subjects with BMIO/WCO phenotype also had a significantly higher risk of NAFLD than males (OR: female 9.78 vs. male 4.19). This suggested that elevated BMI and WC and ABSI may be more important risk factors for NAFLD in females, and females should pay more attention to maintaining normal body weight and body shape to prevent the occurrence of NAFLD.\nOur current study also compared the ability of five continuous combined indicators, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC, to identify NAFLD with BMI and WC and ABSI alone. According to the results of ROC analysis, ARI and OBMI+ABSI had the strongest NAFLD identification ability in both sexes, and although there were only minor differences between the two and OBMI+WC, BMI*WC and BMI*ABSI, they were all significantly higher than BMI and WC and ABSI alone. From Table 5 we found that the optimal proportional combination indicators (OBMI+WC and OBMI+ABSI) have extremely similar diagnostic efficacy to the multiplicative indicators (BMI*WC and BMI*ABSI) and ARI, which was an interesting result. The optimal proportional combination indicators in this study were calculated after directly obtaining the proportional coefficients based on the ratio of the regression coefficients of BMI and WC or ABSI in model 3, which considered the relative weights of the effects of BMI and WC or ABSI on NAFLD and could fully reflect the effects of BMI combined with WC and ABSI on NAFLD. However, the construction of the calculation formula for the optimal proportional combination was relatively complicated, which may be affected by different adjusted confounding factors and changes in the correlation between BMI, WC, ABSI and NAFLD in different test groups. In contrast, the ARI (BMI, ABSI) was obtained by multiplying the logarithms of ORs of BMI and ABSI by the corresponding BMI and ABSI values of each subject and then summing them together [26] and the multiplicative combination indicators were obtained by directly multiplying BMI with WC and ABSI, which are easy to calculate and had a high diagnostic performance in both sexes. In summary, ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be better obesity indicators for clinical screening of NAFLD. Additionally, the results of the ROC analysis after further grouping the sexes by different age groups showed that all combination indicators of BMI and ABSI or WC significantly improved the ability of a single indicator to identify NAFLD in middle-aged females and young and middle-aged males. It is worth mentioning that the highest diagnostic efficacy for NAFLD was found in the young female population for both combined and single obesity indicators; this may be related to the protective effect of large amounts of estrogen on NAFLD in young females. As we all know, the reproductive status of females is closely related to their health status. Experimental studies have shown that estrogens can exert antioxidant, anti-steatosis, and anti-fibrotic effects in the liver [46, 47]. Whereas, a significant decrease in estrogen levels and an increase in circulating androgen levels in postmenopausal females may lead to disorders of lipid metabolism and the development of metabolic syndrome [48]. In addition, the decline in the body’s metabolic status with aging is also an important risk factor for NAFLD [49]. Thus, younger females have a healthier metabolic profile and fewer confounding factors that affect the risk of NAFLD identified by obesity indicators compared with postmenopausal females and males.\nOur findings have important implications for both clinical screening and epidemiological studies of NAFLD. Due to the insidious onset of NAFLD, early-stage patients may be asymptomatic, primary care workers lack attention to NAFLD, and public awareness and treatment rates are very low [50, 51]; moreover, there are no approved treatments for NAFLD, hence the mainstay of prevention and treatment of NAFLD remains healthy lifestyle and weight control [1]. In the current study, ARI and BMI*WC and BMI*ABSI had greater than 80% diagnostic accuracy in all age groups of both sexes and up to 95.22% in young women, and ARI and BMI*WC and BMI*ABSI were easily calculated without additional measurements or equipment requirements. Therefore, we recommend that the ARI and BMI*ABSI and BMI*WC should be incorporated into the clinical screening plan for NAFLD, which will greatly improve the detection rate of NAFLD.\n[SUBTITLE] Advantages and limitations of research [SUBSECTION] Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.\nSeveral advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.","Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.","Taken together, our findings confirmed that BMI combined with ABSI and WC can better explain obesity-related NAFLD risk than a single indicator in the general population, and identified the ARI and the multiplicative combination indicators BMI*ABSI and BMI*WC as the simplest and very effective combination indicators for diagnosing the risk of NAFLD. These new findings underscored the importance of the combined application of BMI and the abdominal obesity index in clinical practice to assess NAFLD risk in the general population and provided a cost-effective tool for precise preventive screening and treatment monitoring for NAFLD.","[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\n\nSupplementary Material 2\n\nSupplementary Material 2\nBelow is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\n\nSupplementary Material 2\n\nSupplementary Material 2","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\n\nSupplementary Material 2\n\nSupplementary Material 2"],"string":"[\n \"Non-alcoholic fatty liver disease (NAFLD) is a chronic non-infectious liver disease characterized by oxidative stress, inflammation, and fibrosis in hepatocytes among people without a history of excessive alcohol consumption [1, 2]. NAFLD can present as asymptomatic simple hepatic steatosis in its early stages, and without intervention, oxidative stress, inflammation, and fibrosis in hepatocytes can contribute to the continuous progression of NAFLD to non-alcoholic steatohepatitis and then to cirrhosis and ultimately to hepatocellular carcinoma [2, 3]. Recent epidemiological surveys have shown that about a quarter of the global population suffers from NAFLD, with a 27.4% estimated prevalence in Asia [4]. NAFLD is not only the most common chronic liver disease worldwide but has replaced hepatitis B virus infection as the primary risk factor for advanced liver diseases such as cirrhosis and hepatocellular carcinoma [5]. The high prevalence of NAFLD and its severe complications place a heavy burden on humans and the world’s health care systems, making the early prevention, identification, and intervention of NAFLD of great public health importance.\\nObesity is one of the most important risk factors for NAFLD [6]. Waist circumference (WC) and body mass index (BMI) are currently the most widely used basic body measures for the assessment of general and central obesity and also are known risk factors for NAFLD [7–10]. Both BMI and WC alone, however, have some distinct limitations, such as the inability of BMI alone to distinguish between adipose and muscle tissue and the inability of WC to determine the differential contribution of abdominal visceral adipose tissue and abdominal subcutaneous adipose tissue to central obesity [11–13]. It is not clear if combining risk associated with BMI and WC can improve the individual risk association. Several relevant studies have analyzed the performance of the combination of BMI and WC in disease risk assessment, and overall, combining BMI and WC in different ways significantly improved the predictive efficacy for obesity-related diseases such as obesity-related hypertension, type 2 diabetes, cardiovascular disease, and all-cause mortality risk compared to a single indicator [14–17]. In addition, a recent study by Wang et al. also found that high levels of WC and BMI were both associated with NAFLD risk in people with type 2 diabetes and that BMI combined with WC was superior to the single measures in assessing the risk of NAFLD [18]. What remains unclear, however, is whether the combination of BMI and WC best identifies a higher risk of NAFLD in the general population. Furthermore, it is important to note that there is a strong correlation between BMI and WC (r ≈ 0.80), and directly combining BMI with WC has the potential to confound the association between BMI and NAFLD, leading to biased and possibly misleading risk estimates [19–21]. A Body Shape Index (ABSI), an allometric abdominal obesity index calculated from WC, height, and weight, is intended to be independent of BMI and has proven to be useful for identifying individuals at risk for sarcopenic obesity [22]; therefore, we also evaluated combining BMI with the non-correlated abdominal obesity index ABSI. This study sought to explore risk assessment for NAFLD of various combinations of BMI and WC or ABSI in a cohort of general health examinees from NAGALA.\",\n \"[SUBTITLE] Data sources and study design [SUBSECTION] In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\\nIn the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\\n[SUBTITLE] Definition and derivation of the combined indicators of BMI with WC and ABSI [SUBSECTION] Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\\nBased on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\\n[SUBTITLE] Combination using usual cut-offs for BMI and WC [SUBSECTION] In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\\nIn accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\\n[SUBTITLE] Optimal proportional combinations OBMI+WC and OBMI+ABSI, multiplicative combination indicators BMI*ABSI and BMI*WC, and ARI [SUBSECTION] The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\\nThe optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\\n[SUBTITLE] Diagnosis of NAFLD [SUBSECTION] Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\\nSpecialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\\n[SUBTITLE] Statistical analysis [SUBSECTION] We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\\nWe performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\",\n \"In the current study, we used data from 14,251 medical examiners recruited by the NAGALA cohort between 1994 and 2016 [23]. NAGALA is an ongoing longitudinal cohort study initiated by Murakami Memorial Hospital in 1994, using physical examination data from the general population to assess common risk factors for the development of chronic diseases; a detailed description of the NAGALA cohort study design has been previously published and the study data have been uploaded by Professor Okamura to the Dryad database for public access [24]. For the current study, we excluded from the original data set (1) subjects who were diagnosed or self-reported with hepatitis (viral/alcoholic) or diabetes and impaired fasting glucose at baseline (n = 1,547); (2) subjects with alcohol abuse (n = 1,952, males consuming ≥ 210 g of alcohol per week and females consuming ≥ 140 g) [25]; (3) subjects taking any medications at baseline (n = 2,321); and, (4) subjects with missing anthropometric and medical examination data (n = 873) were also excluded. We ultimately included 14,251 eligible subjects, with the inclusion and exclusion criteria shown in Fig. 1. The NAGALA cohort study has been ethically reviewed by the Murakami Memorial Hospital Institutional Ethics Review Committee (IRB2018-09-01) and all subjects signed written informed consent for the use of their data; The current study was a post-hoc analysis of the NAGALA cohort, and the Jiangxi Provincial People’s Hospital Ethics Review Committee reviewed the study design and granted approval (IRB2021-066).\\n\\nFig. 1Flowchart of the selection process of study subjects\\n\\nFlowchart of the selection process of study subjects\\nIn this study, we extracted the following data from each subject: anthropometric and demographic information including weight, height, sex, age, WC, BMI, ABSI, smoking and drinking status, and exercise habits of the subjects; where drinking status was classified according to the subject’s weekly alcohol consumption in the previous month as moderate drinking, light drinking and never or small drinking; smoking status was defined as never, previous and current smoking; for exercise habits, subjects were considered to have exercise habits if they had at least one physical activity of any form per week. BMI was calculated as weight (kg)/[height (m)]2 and ABSI was calculated as 1,000*WC (m)*[weight (kg)] – 2/3*[height (m)]5/6 [22].\\nData on laboratory tests, including blood pressure, triglycerides (TG), aspartate aminotransferase (AST), total cholesterol (TC), fasting blood glucose (FPG), gamma-glutamyl transferase (GGT), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), and glycated hemoglobin (HbA1c). The specific steps for obtaining demographic and anthropometric information and data from laboratory tests have been described in detail elsewhere [19]; all procedures in this study fully followed the Declaration of Helsinki. See STROBE statement (S1 Text).\",\n \"Based on previous studies, the present study included four combination types of BMI with WC and ABSI: (1) Using the WHO-recommended clinical cut-offs of WC and BMI combined into four obesity phenotypes; (2) constructing the optimal proportional combination of BMI with WC and ABSI, OBMI+WC and OBMI+ABSI; and (3) using the multiplicative combination of WC and ABSI with BMI, BMI*WC, and BMI*ABSI; (4) construction of anthropometric risk index [ARI (BMI, ABSI)] for combined risk of BMI and ABSI using the method recommended by Krakauer NY et al. [26].\",\n \"In accordance with the World Health Organization (WHO) expert committee cut-off recommendations for Asian populations, we defined BMI ≥ 25 kg/m2 as overweight/obese and BMI < 25 kg/m2 as normal weight [27]. We defined WC ≥ 85 cm for males or WC ≥ 80 cm for females as central obesity and WC < 85 cm for males or WC < 80 cm for females as normal WC [28]. All subjects were assigned to the following four obesity phenotypes according to the baseline WC and BMI: (1) BMIN/WCN: normal weight + normal WC; (2) BMIO/WCN: overweight/obese + normal WC; (3) BMIN/WCO: normal weight + central obesity; (4) BMIO/WCO: overweight/obese + central obesity.\",\n \"The optimal proportional combination indicator OBMI+WC was calculated as OBMI+WC= nWC*WC + nBMI*BMI. The optimal proportion coefficients for WC (nWC) and BMI (nBMI) were calculated as nWC = βWC / (βBMI + βWC) and nBMI = βBMI / (βBMI + βWC), respectively; where βWC and βBMI were the regression coefficients of WC and BMI in the multivariable logistic regression model, respectively. To acquire regression coefficients for WC and BMI, we included the presence or absence of NAFLD as the outcome event of interest, WC and BMI as variables in a multivariate-adjusted logistic regression model, and age, height, ALT, AST, systolic blood pressure (SBP), TG, GGT, HDL-C, HbA1c, TC, FPG, and drinking status were adjusted as covariates. Similarly, we used the same method to obtain OBMI+ABSI. Since NAFLD risk was positively correlated with BMI, WC, and ABSI, and there was no correlation between BMI and ABSI, which is consistent with the construction of an ARI (BMI, ABSI) for the combined risk of BMI and ABSI, and we calculated ARI (BMI, ABSI) using the method provided by Krakauer NY et al. [26]. In addition, we calculated the products BMI*WC and BMI* ABSI for further analysis.\",\n \"Specialized gastroenterologists, blinded to the subjects scored each abdominal ultrasound for NAFLD risk based on four criteria: (1) changes in the intensity of liver blood flow signals (0–4 points); (2) the clarity of liver blood vessels (0–1 points); (3) liver and kidney echo contrast (0–4 points); (4) deep liver echo attenuation (0–2 points). NAFLD was diagnosed if the sum of the above four scores was greater than 2 [29].\",\n \"We performed all statistical analysis steps in the current study using Empower (R) version 2.2 and R language version 3.4.3. Significance was defined as P < 0.05 in all comparisons (two-sided). Given the significant differences in body composition, fat deposition patterns, and prevalence of NAFLD between males and females [30], all analyses in the present study were performed separately for female and male subjects and the specific steps were as follows:\\nFirst, we calculated the quartile categories of BMI*ABSI using the quartile function and subsequently described baseline information by grouping all subjects based on the quartile categories of BMI*ABSI and whether they had NAFLD. For the measurement data, QQ plots were first used to determine the pattern of data distribution, and the data with a normal and a skewed distribution were described as mean (standard deviation) and median (interquartile range), respectively. T-test or one-way ANOVA was used for group comparisons of normal data, and Mann-Whitney rank-sum test or Kruskal-Wallis rank-sum test was used for group comparisons of skewed data. Count data were described as frequencies (%), using the Pearson chi-square test for comparison between groups.\\nSecond, before exploring the association of each obesity indicator with NAFLD, we first analyzed the correlations between several anthropometric measures (height, BMI, WC, ABSI), and it is worth mentioning that a combination of two basic anthropometric indicators of obesity which have smaller correlations is considered more appropriate. Additionally, we also performed a collinearity screening for all covariates [31], where the weight and diastolic blood pressure (DBP) had variance inflation factors (VIF) greater than 5 and were therefore treated as collinear variables and not included in the subsequent model adjustment (Supplementary Table 1). We developed one univariate and three multivariable logistic regression models for assessing the associations between the continuous variables BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and the categorical variable obesity phenotypes and risk of NAFLD, and recorded the corresponding Odds ratio (OR) and 95% confidence interval (CI), respectively. Due to the large numerical span of BMI*WC (689.56-6490.01) and BMI*ABSI (961.13-3630.94), the change in effect value caused by the change of one unit was small, so the BMI*WC and BMI*ABSI were Z-score transformed and included in the regression models. In the model adjustment of multivariable logistic regression, model 1 considered the effect of anthropometric parameters of age and height on NAFLD; model 2 adjusted drinking status and important lipid parameters (TC, TG, HDL-C) on the basis of model 1; model 3 further considered the effects of liver enzyme indicators (ALT, AST, GGT), blood glucose and blood pressure parameters (FPG, HbA1c, SBP) on the basis of model 2. In addition, to further explore the correlation between BMI*ABSI and NAFLD, we also included BMI*ABSI as a categorical variable in the logistic regression models and calculated the trend of the association between the median of each category of BMI*ABSI and NAFLD.\\nThird, based on the results of the correlation analysis between each obesity indicator and NAFLD, we also constructed receiver operating characteristic (ROC) curves to evaluate the ability of continuous obesity indicator BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC to identify the risk of NAFLD. The area under the curve (AUC) and Delong test were used to compare the differences in the ability to identify NAFLD risk of each obesity indicator in both sexes and different age subgroups of both sexes, and the optimal diagnostic threshold for each indicator was calculated using Youden’s index.\",\n \"[SUBTITLE] Demographic and clinical characteristics of subjects [SUBSECTION] The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\\n\\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\n\\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\\n\\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\\n\\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\\nAbbreviations as in Table 1\\nThe study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\\n\\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\n\\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\\n\\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\\n\\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\\nAbbreviations as in Table 1\\n[SUBTITLE] Calculation of the optimal proportional combination of BMI with WC and ABSI [SUBSECTION] In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\\n\\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\\n\\nOptimal combination equations of BMI with WC and ABSI.\\nAbbreviations as in Table 1\\nIn the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\\n\\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\\n\\nOptimal combination equations of BMI with WC and ABSI.\\nAbbreviations as in Table 1\\n[SUBTITLE] Association of various obesity indicators with NAFLD [SUBSECTION] Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\\n\\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\n\\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\\nModel 1 adjusted for age and height\\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\nCorrelation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\\n\\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\n\\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\\nModel 1 adjusted for age and height\\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\n[SUBTITLE] Accuracy of various obesity indicators to identify NAFLD [SUBSECTION] ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\\n\\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\\n\\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\\nABSI**\\nARI**\\n0.6171\\n0.8912\\n0.5916\\n0.8775\\n0.6426\\n0.9048\\n75.7925\\n12.7825\\n0.5926\\n0.7735\\n0.5962\\n0.8577\\n0.1888\\n0.6312\\nAbbreviations as in Table 1\\n**P < 0.05 compared with BMI*WC in each gender\\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\\n\\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\\n\\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\\nAbbreviations as in Table 1\\n***P < 0.05 compared with BMI*WC in each age group of both genders\\nROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\\n\\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\\n\\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\\nABSI**\\nARI**\\n0.6171\\n0.8912\\n0.5916\\n0.8775\\n0.6426\\n0.9048\\n75.7925\\n12.7825\\n0.5926\\n0.7735\\n0.5962\\n0.8577\\n0.1888\\n0.6312\\nAbbreviations as in Table 1\\n**P < 0.05 compared with BMI*WC in each gender\\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\\n\\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\\n\\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\\nAbbreviations as in Table 1\\n***P < 0.05 compared with BMI*WC in each age group of both genders\",\n \"The study included 14,251 subjects, of whom 7,411 (52%) were males, with a mean age of 43.82 (8.99) years, and 2,029 (27.37%) were diagnosed with NAFLD; 6,840 were (48%) females, with mean age 43.22 (8.78) years, and 478 (6.99%) were diagnosed with NAFLD. Table 1 is stratified by gender and describes the differences in baseline characteristics between NAFLD patients and non-NAFLD subjects. We found that NAFLD patients consistently had higher age, weight, TC, SBP, DBP, HbA1c, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, ALT, AST, GGT, FPG, and TG levels in both sexes, but with lower HDL-C levels. Notably, non-NAFLD subjects of both sexes tended to have greater alcohol consumption, which paradoxically appeared to be a protective factor for NAFLD. Additionally, the comparison between the NAFLD group and the non-NAFLD group showed that physical exercise habits were only significantly different in males, height was only significantly different in females, and smoking habits were not significantly different in both sexes.\\n\\nTable 1Characteristics of 14,251 subjects stratified by NAFLD and genderMaleFemaleVariablesNon-NAFLDn = 5382 (72.61%)NAFLDn = 2029 (27.39%)P-valueNon-NAFLDn = 6362 (93.01%)NAFLDn = 478 (6.99%)P-valueAge, years43.7 (9.3)44.1 (8.2)0.00242.9 (8.7)47.6 (8.3)< 0.001Height, cm170.9 (6.0)170.6 (5.9)0.084158.4 (5.4)157.0 (5.3)< 0.001Weight, kg64.6 (8.3)74.3 (10.6)< 0.00151.9 (7.1)63.2 (10.0)< 0.001TC, mmol/L5.1 (0.8)5.4 (0.9)< 0.0015.1 (0.9)5.6 (0.9)< 0.001HDL-C, mmol/L1.3 (0.3)1.1 (0.3)< 0.0011.7 (0.4)1.4 (0.3)< 0.001SBP, mmHg116.0 (13.2)124.0 (14.5)< 0.001108.4 (13.8)120.7 (16.0)< 0.001FPG, mmol/L5.3 (0.4)5.4 (0.3)< 0.0015.0 (0.4)5.3 (0.4)< 0.001DBP, mmHg72.9 (9.3)78.4 (10.1)< 0.00167.0 (9.5)75.1 (10.2)< 0.001HbA1c, %5.1 (0.3)5.3 (0.3)< 0.0015.2 (0.3)5.4 (0.3)< 0.001BMI, kg/m222.1 (2.4)25.5 (3.0)< 0.00120.7 (2.6)25.6 (3.6)< 0.001WC, cm78.0 (6.8)86.6 (7.4)< 0.00170.8 (7.3)83.3 (8.9)< 0.001ABSI75.8 (3.3)76.7 (3.1)< 0.00174.8 (4.6)76.8 (4.5)< 0.001ARI13.1 (0.9)14.3 (1.0)< 0.00112.2 (1.0)14.0 (1.3)< 0.001OBMI+WC34.4 (3.2)38.9 (3.8)< 0.00131.7 (3.4)38.3 (4.5)< 0.001OBMI+ABSI32.0 (2.1)34.9 (2.4)< 0.00128.9 (2.3)33.4 (3.0)< 0.001BMI*WC1714.2 (1507.7-1939.1)2152.9 (1919.2-2457.9)< 0.0011423.8 (1249.1-1645.8)2070.2 (1795.3-2396.7)< 0.001BMI*ABSI1669.3 (1538.1–1806.0)1922.5 (1795.9-2082.3)< 0.0011518.4 (1390.8-1673.5)1917.8 (1756.7-2131.4)< 0.001ALT, U/L18.0 (14.0–23.0)29.0 (22.0–41.0)< 0.00113.0 (11.0–17.0)19.0 (15.0–26.0)< 0.001AST, U/L17.0 (14.0–21.0)21.0 (17.0–26.0)< 0.00116.0 (13.0–19.0)18.0 (15.0–22.0)< 0.001GGT, U/L17.0 (14.0–24.0)24.0 (18.0–35.0)< 0.00112.0 (9.0–14.0)15.0 (12.0–20.0)< 0.001TG, mmol/L0.8 (0.6–1.2)1.3 (0.9–1.9)< 0.0010.5 (0.4–0.8)1.0 (0.7–1.4)< 0.001Habit of exercise, No. (%)< 0.0010.335No4300 (79.9%)1720 (84.8%)5351 (84.1%)410 (85.8%)Yes1082 (20.1%)309 (15.2%)1011 (15.9%)68 (14.2%)Drinking status, No. (%)< 0.0010.004Non or small3731 (69.3%)1623 (80.0%)5986 (94.1%)465 (97.3%)Light1096 (20.4%)273 (13.5%)376 (5.9%)13 (2.7%)Moderate555 (10.3%)133 (6.6%)Smoking status, No. (%)0.0670.664Non1952 (36.3%)758 (37.4%)5609 (88.2%)427 (89.3%)Past1538 (28.6%)615 (30.3%)382 (6.0%)24 (5.0%)Current1892 (35.2%)656 (32.3%)371 (5.8%)27 (5.6%)Values were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\n\\nCharacteristics of 14,251 subjects stratified by NAFLD and gender\\nValues were expressed as mean (standard deviation) or medians (quartile interval) or n (%). Abbreviations: BMI: body mass index; WC: waist circumference; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transferase; HDL-C: high-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; FPG: fasting plasma glucose; HbA1c: hemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; NAFLD: non-alcoholic fatty liver disease; ABSI: A body shape index; OBMI+ABSI: Optimal proportional combination of BMI and ABSI; OBMI+WC: Optimal proportional combination of BMI and WC; ARI: Anthropometric risk index\\nBaseline characteristics for all subjects according to the quartile category of BMI*ABSI are shown in Table 2. We found that with the increase of BMI*ABSI quartiles, all baseline metrics of subjects showed a trend of change (all P < 0.001). Of these, a greater proportion of male subjects were in the higher BMI*ABSI quartiles, with a male-to-female ratio of almost 7:3 in Q4, and there were more subjects with smoking and drinking habits and with less physical exercise in Q4. Moreover, subjects’ HDL-C levels decreased with increasing BMI*ABSI quartiles, while age, BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*WC, height, weight, TC, SBP, HbA1c, ALT, AST, DBP, FPG, GGT, TG levels, and the prevalence of NAFLD increased (Q4: 48.5%).\\n\\nTable 2Characteristics of 14,251 subjects stratified according to the BMI*ABSI quartilesBMI*ABSI quartilesvariablesQ1 (961.1-1477.2)Q2 (1477.4-1645.5)Q3 (1645.6-1827.4)Q4 (1827.5-3630.9)P-valueNo. of subjects3563356235633563Male, No. (%)867 (24.3%)1666 (46.8%)2269 (63.7%)2609 (73.2%)Age, years41.0 (8.7)43.1 (8.7)44.5 (8.6)45.6 (8.9)< 0.001BMI, kg/m218.7 (1.3)20.9 (1.2)22.7 (1.2)26.0 (2.5)< 0.001WC, cm65.4 (3.7)72.8 (2.9)78.8 (2.9)87.7 (5.8)< 0.001ABSI73.1 (3.8)75.1 (3.8)76.5 (3.5)77.5 (3.5)< 0.001ARI11.4 (0.4)12.4 (0.3)13.2 (0.3)14.4 (0.8)< 0.001OBMI+WC29.0 (1.5)32.3 (1.0)35.0 (1.1)39.6 (3.0)< 0.001OBMI+ABSI27.4 (1.3)29.9 (1.1)31.9 (1.1)35.0 (2.0)< 0.001BMI*WC1241.2 (1147.1-1318.4)1514.2 (1447.2-1587.4)1782.6 (1703.2-1865.8)2194.4 (2051.7-2436.9)< 0.001Height, cm161.7 (7.7)164.3 (8.3)166.1 (8.5)167.1 (8.4)< 0.001Weight, kg49.0 (5.7)56.4 (6.5)62.7 (7.3)72.9 (10.3)< 0.001TC, mmol/L4.9 (0.8)5.0 (0.9)5.2 (0.9)5.4 (0.9)< 0.001HDL-C, mmol/L1.7 (0.39)1.5 (0.4)1.4 (0.4)1.2 (0.3)< 0.001SBP, mmHg105.6 (12.5)111.1 (12.8)115.9 (13.3)123.1 (14.8)< 0.001FPG, mmol/L4.9 (0.4)5.1 (0.4)5.2 (0.4)5.3 (0.4)< 0.001DBP, mmHg65.5 (8.5)69.0 (9.1)72.5 (9.6)77.6 (10.2)< 0.001HbA1c, %5.1 (0.3)5.1 (0.3)5.2 (0.3)5.3 (0.3)< 0.001ALT, U/L14.0 (11.0–17.0)15.0 (12.0–19.0)18.0 (13.0–23.0)23.0 (17.0–33.0)< 0.001AST, U/L16.0 (13.0–19.0)16.00 (14.0–20.0)17.0 (14.0–21.0)19.0 (16.0–24.0)< 0.001GGT, U/L12.0 (10.0–15.0)13.00 (11.0–18.0)16.0 (12.0–22.0)21.0 (15.0–30.0)< 0.001TG, mmol/L0.51 (0.4–0.7)0.6 (0.5–0.9)0.8 (0.6–1.2)1.1 (0.8–1.6)< 0.001Habit of exercise, No. (%)< 0.001No2963 (83.2%)2858 (80.2%)2920 (82.0%)3040 (85.3%)Yes600 (16.8%)704 (19.8%)643 (18.1%)523 (14.7%)Drinking status, No. (%)< 0.001Non or small3228 (90.6%)2937 (82.5%)2848 (79.9%)2792 (78.4%)Light279 (7.8%)467 (13.1%)507 (14.2%)505 (14.2%)Moderate56 (1.6%)158 (4.4%)208 (5.8%)266 (7.5%)Smoking status, No. (%)< 0.001Non2736 (76.8%)2322 (65.2%)1957 (54.9%)1731 (48.6%)Past337 (9.5%)560 (15.7%)785 (22.0%)877 (24.6%)Current490 (13.8%)680 (19.1%)821 (23.0%)955 (26.8%)NAFLD, No. (%)16 (0.5%)150 (4.2%)611 (17.2%)1730 (48.5%)< 0.001Abbreviations as in Table 1\\n\\nCharacteristics of 14,251 subjects stratified according to the BMI*ABSI quartiles\\nAbbreviations as in Table 1\",\n \"In the logistic regression equation constructed to calculate the optimal proportional combination OBMI+WC, the regression coefficient βBMI for BMI was 0.218 (0.154–0.282), and βWC for WC was 0.063 (0.038–0.089) in females; in males, βBMI was 0.177 (0.124–0.299), and βWC was 0.068 (0.048–0.089). Therefore, according to the calculation formula of the optimal proportional coefficients, the values of nBMI and nWC in females were 0.776 and 0.224, respectively, and the values of nBMI and nWC in males were 0.722 and 0.278, respectively. Similarly, in the logistic regression equation constructed to calculate the optimal proportional combination OBMI+ABSI, βBMI and βABSI were 0.356 and 0.064, respectively, in females, resulting in nBMI and nABSI values of 0.848 and 0.152, respectively; in males, βBMI and βABSI were 0.334 and 0.075, respectively, and nBMI and nABSI values were 0.817 and 0.183, respectively. Table 3 shows the calculation formulas of OBMI+WC and OBMI+ABSI.\\n\\nTable 3Optimal combination equations of BMI with WC and ABSI.Logistic regression derived equationsMaleFemaleequations for BMI and WC0.722BMI + 0.278WC0.776BMI + 0.224WCequations for BMI and ABSI0.817BMI + 0.183ABSI0.848BMI + 0.152ABSIAbbreviations as in Table 1\\n\\nOptimal combination equations of BMI with WC and ABSI.\\nAbbreviations as in Table 1\",\n \"Correlation analysis of the basic anthropometric measures showed a strong correlation between BMI and WC in both sexes (r: male 0.8796; female 0.8191), while BMI was barely correlated with ABSI (r < 0.05) (Supplementary Table 2). In contrast, the combination of ABSI and BMI may be more appropriate. Table 4 presents the associations between BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC and the risk of NAFLD in both sexes. In multivariate-adjusted logistic regression models, BMI and WC, and ABSI were positively correlated with NAFLD risk. In the current study, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC were positively correlated with NAFLD risk in all four logistic regression models, and the OR values in the univariate logistic regression model ranged from 1.44 to 5.02; after further adjustment for age, height, GGT, TC, AST, HDL-C, drinking status, ALT, TG, FPG, HbA1c, and SBP (Model 3), the results did not change significantly. The risk of NAFLD increased by 33% (OR 1.33, 95%CI 1.29, 1.37) and 52% (OR 1.52, 95%CI 1.45, 1.60) and 172% (OR 2.72, 95%CI 2.43, 3.04), in females, for each unit increase in OBMI+WC and OBMI+ABSI and ARI respectively. Similarly, in males, NAFLD risk was increased by 30% (OR 1.30, 95%CI 1.27, 1.33) and 50% (OR 1.50, 95%CI 1.45, 1.56) and 172% (OR 2.72, 95%CI 2.48, 2.98), respectively. Each standard deviation increase in BMI*WC and BMI*ABSI in females was associated with a 213% (OR 3.13, 95%CI 2.74,3,56) and 209% (OR 3.09, 95%CI 2.72, 3.52) increased risk for NAFLD respectively, and for males 185% (OR 2.85, 95%CI 2.59, 3.15) and 197% (OR 2.97, 95%CI 2.68, 3.28), respectively. Furthermore, the association between BMI*ABSI and NAFLD remained unchanged after treating BMI*ABSI as a categorical variable; taking the first quartile as the control group, the NAFLD risk increased with the increase of BMI*ABSI quartiles (all P for trend < 0.001). Moreover, we also analyzed the risk of NAFLD for each obesity phenotype in multivariable logistic regression models (Supplementary Table 3), which in model 3 showed that compared with the BMIN/WCN phenotype, male and female BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes all had a significantly higher risk of NAFLD. The risk of NAFLD associated with the BMIO/ WCN, BMIN/WCO, and BMIO/WCO phenotypes had ORs of 3.67, 3.77, and 9.78 in females and 2.76, 2.74, and 4.19 in males, respectively. Clearly, subjects with the BMIO/WCO phenotype consistently had the highest risk of developing NAFLD.\\n\\nTable 4Odds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjectsOdds ratios (95% confidence interval)Crude modelModel 1Model 2Model 3FemaleBMI1.58 (1.53, 1.64)*1.58 (1.52, 1.63)*1.49 (1.44, 1.55)*1.42 (1.36, 1.47)*WC1.19 (1.17, 1.21)*1.19 (1.17, 1.21)*1.17 (1.15, 1.18)*1.14 (1.12, 1.16)*ABSI1.09 (1.07, 1.12)*1.07 (1.05, 1.09)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI3.70 (3.36, 4.07)*3.61 (3.28, 3.98)*3.11 (2.80, 3.44)*2.72 (2.43, 3.04)*OBMI+ABSI1.73 (1.66, 1.80)*1.71 (1.65, 1.79)*1.61 (1.54, 1.68)*1.52 (1.45, 1.60)*OBMI+WC1.44 (1.41, 1.48)*1.44 (1.40, 1.48)*1.38 (1.34, 1.42)*1.33 (1.29, 1.37)*BMI*ABSI (Per SD)4.49 (4.02, 5.02)*4.33 (3.87, 4.84)*3.63 (3.23, 4.09)*3.09 (2.72, 3.52)*BMI*WC (Per SD)4.50 (4.03, 5.03)*4.41 (3.94, 4.93)*3.68 (3.27, 4.15)*3.13 (2.74, 3.56)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 210.90 (4.64, 25.59)*10.32 (4.39, 24.24)*8.42 (3.57, 19.86)*7.47 (3.16, 17.65)*Quartile 343.73 (19.20, 99.63)*39.09 (17.13, 89.20)*25.14 (10.94, 57.76)*19.49 (8.44, 44.98)*Quartile 4217.88 (96.72, 490.84)*188.51 (83.50, 425.59)*105.60 (46.43, 240.20)*63.97 (27.87, 146.85)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001MaleBMI1.62 (1.58, 1.67)*1.62 (1.58, 1.67)*1.52 (1.48, 1.57)*1.38 (1.34, 1.42)*WC1.19 (1.18, 1.20)*1.21 (1.20, 1.22)*1.18 (1.17, 1.19)*1.13 (1.12, 1.15)*ABSI1.09 (1.07, 1.10)*1.10 (1.08, 1.12)*1.06 (1.04, 1.08)*1.05 (1.02, 1.07)*ARI4.39 (4.06, 4.75)*4.40 (4.07, 4.77)*3.68 (3.38, 4.00)*2.72 (2.48, 2.98)*OBMI+ABSI1.83 (1.77, 1.89)*1.83 (1.77, 1.89)*1.70 (1.65, 1.76)*1.50 (1.45, 1.56)*OBMI+WC1.47 (1.44, 1.50)*1.47 (1.44, 1.50)*1.41 (1.38, 1.44)*1.30 (1.27, 1.33)*BMI*ABSI (Per SD)5.02 (4.61, 5.48)*5.08 (4.66, 5.55)*4.16 (3.80, 4.56)*2.97 (2.68, 3.28)*BMI*WC (Per SD)4.72 (4.34, 5.13)*4.83 (4.44, 5.26)*3.95 (3.62, 4.32)*2.85 (2.59, 3.15)*BMI*ABSI (quartiles)Quartile 11.01.01.01.0Quartile 25.76 (3.00, 11.09)*5.89 (3.06, 11.33)*4.68 (2.42, 9.02)*3.97 (2.03, 7.78)*Quartile 323.97 (12.75, 45.07)*25.00 (13.29, 47.03)*15.83 (8.38, 29.90)*10.27 (5.35, 19.72)*Quartile 4102.03 (54.44, 191.23)*108.71 (57.95, 203.93)*58.44 (30.99, 110.21)*26.15 (13.61, 50.25)*P for trend< 0.0001< 0.0001< 0.0001< 0.0001*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1Model 1 adjusted for age and heightModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking statusModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\\n\\nOdds ratios and 95% confidence intervals of BMI, WC, ABSI, OBMI+WC, OBMI+ABSI, BMI*ABSI, BMI*WC, and ARI for NAFLD in female and male subjects\\n*P < 0.0001; Abbreviation: SD: standard deviation; other abbreviations as in Table 1\\nModel 1 adjusted for age and height\\nModel 2 adjusted for model 1 plus TC, TG, HDL-C, drinking status\\nModel 3 adjusted for model 2 plus ALT, AST, GGT, FPG, HbA1c, and SBP.\",\n \"ROC curves were drawn to assess the accuracy of BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC for identifying NAFLD in both sexes (Fig. 2). Table 5 summarizes the AUC and optimal diagnostic thresholds for these obesity indicators. In contrast, ARI and OBMI+ABSI had the same and highest AUC in both female and male subjects, with AUC and optimal diagnostic thresholds of 0.8270, 33.2569 and 0.8912, 30.4143 for ARI in males and females, respectively. The Delong test further showed that both the combined indicators of BMI and WC (OBMI+WC and BMI*WC) and BMI and ABSI (OBMI+ABSI and BMI*ABSI and ARI) were significantly better at identifying the risk of NAFLD than BMI or WC or ABSI alone (Delong test P < 0.05).\\n\\nFig. 2ROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nROC curve analysis of BMI, WC, ABSI, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC in females and males\\n\\nTable 5The best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleBMI**0.87990.86480.895022.75650.81770.79500.6127WC**0.86950.85450.884474.25000.70500.87450.5804ABSI**ARI**0.61710.89120.59160.87750.64260.904875.792512.78250.59260.77350.59620.85770.18880.6312OBMI+ABSI**0.89120.87750.904830.41430.77330.85770.6310OBMI+WC0.88880.87500.902734.12340.77160.85560.6272BMI*ABSI0.88840.87480.90211704.21000.79220.82850.6207BMI*WC0.88880.87500.90271674.11830.77330.85560.6286MaleBMI**0.81600.80550.826423.55550.73820.73090.4691WC**0.81020.79980.820780.65000.66740.80340.4708ABSI**0.57950.56530.593674.91280.40520.72400.1292ARI**0.82700.81700.836913.60960.74560.74570.4913OBMI+ABSI**0.82700.81700.836933.25690.74580.74570.4915OBMI+WC0.82420.81410.834339.83130.74380.74470.4885BMI*ABSI0.82570.81570.83561758.75900.67610.81170.4878BMI*WC0.82440.81430.83451944.04330.75510.73290.4880Abbreviations as in Table 1**P < 0.05 compared with BMI*WC in each gender\\n\\nThe best threshold, sensitivities, specificities, Youden’s index, and area under the curve of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI in different genders\\nABSI**\\nARI**\\n0.6171\\n0.8912\\n0.5916\\n0.8775\\n0.6426\\n0.9048\\n75.7925\\n12.7825\\n0.5926\\n0.7735\\n0.5962\\n0.8577\\n0.1888\\n0.6312\\nAbbreviations as in Table 1\\n**P < 0.05 compared with BMI*WC in each gender\\nTo explore the changes in BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC on the risk identification ability of NAFLD in different age subgroups of both sexes, we constructed ROC curves for 20–39, 40–59, and ≥ 60 years in both sexes, respectively; and the corresponding AUC and optimal diagnostic thresholds were summarized in Table 6. We found that ARI and OBMI+ABSI exhibited considerably higher AUCs in all age subgroups for both sexes except in the female ≥ 60-year age group; the AUC of OBMI+ABSI in the female 20–39 age group was 0.9523, the highest value among all age subgroups of both sexes. In the male 20-59-year-old group and the female 40-59-year-old group, the AUC values of ARI and OBMI+ABSI were slightly higher than those of BMI*WC, but all were significantly higher than those of BMI or WC or ABSI alone, while in the female 20–39 years group and the female ≥ 60 years group AUC values for all combined obesity indicators were significantly higher than that of WC and ABSI; in addition, in the male ≥ 60-year-old group, the seven indicators of obesity had similar NAFLD identification abilities, only ABSI underperformed. Therefore, the combined indicators ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, and OBMI+WC can significantly improve the ability of simple obesity parameters BMI and WC and ABSI to identify NAFLD in middle-aged females and young and middle-aged males and had the highest diagnostic performance in young females.\\n\\nTable 6The area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both gendersAUC95%CI low95%CI upBest thresholdSpecificitySensitivityYouden’s indexFemaleAge 20–39 yearsBMI0.95110.93360.968523.12660.89810.87230.7704WC***0.93190.90790.955976.60000.85600.88300.7390ABSI***0.56860.50850.628875.57420.61220.53190.1441ARI0.95220.93370.970812.79220.84980.93620.7860BMI*WC0.95160.93310.97011675.47840.84360.92550.7691BMI*ABSI0.94860.92890.96821701.34860.86140.92550.7869OBMI+ABSI0.95230.93370.970830.43680.84980.93620.7860OBMI+WC0.95160.93300.970134.12960.84240.92550.7679Age 40–59 yearsBMI***0.84840.82820.868622.75270.78410.77440.5585WC***0.84230.82300.861674.25000.66300.86910.5321ABSI***0.61860.58970.647675.79250.58080.60720.1880ARI***0.86220.84380.880513.02830.80770.77440.5821BMI*WC0.86030.84170.87881674.11830.73320.83840.5716BMI*ABSI0.86020.84200.87851704.21000.75460.81340.5680OBMI+ABSI***0.86220.84380.880530.99950.80790.77440.5823OBMI+WC0.86020.84170.878835.03630.80930.76320.5725Age ≥ 60 yearsBMI0.83080.74580.915923.33340.84390.76000.6039WC***0.76960.67780.861379.75000.77070.68000.4507ABSI***0.56600.45510.677078.18360.50240.72000.2224ARI0.81950.73910.899912.92520.70240.84000.5424BMI*WC0.81990.73960.90021686.12800.69760.84000.5376BMI*ABSI0.80140.71860.88411659.09940.56100.96000.5210OBMI+ABSI0.81950.73920.899830.75250.70240.84000.5424OBMI+WC0.82010.73930.900934.33930.70240.84000.5424MaleAge 20–39 yearsBMI***0.84390.82750.860223.47820.74950.76650.5160WC***0.84150.82550.857580.45000.72190.80090.5228ABSI***0.60630.58260.630074.50900.48710.69910.1862ARI***0.85490.83960.870213.50260.76000.78440.5444BMI*WC0.85130.83570.86691862.54150.72190.81140.5333BMI*ABSI***0.85520.83990.87041766.28860.75570.78740.5431OBMI+ABSI***0.85490.83960.870233.00760.76140.78290.5443OBMI+WC0.85140.83580.866939.14530.72900.80840.5374Age 40–59 yearsBMI***0.80070.78670.814723.25050.67790.76390.4418WC***0.79080.77660.805181.25000.65830.77950.4378ABSI***0.56150.54290.580175.42940.38970.70050.0902ARI***0.81040.79690.824013.61420.72320.74510.4683BMI*WC0.80800.79430.82171955.12030.74480.71770.4625BMI*ABSI0.80860.79500.82221816.10160.73340.72870.4621OBMI+ABSI***0.81050.79690.824033.27330.72390.74430.4682OBMI+WC0.80750.79380.821240.11920.74750.71850.4660Age ≥ 60 yearsBMI0.79030.73630.844423.74200.81110.69510.5062WC0.78040.72820.832680.50000.59440.86590.4603ABSI***0.58040.51190.649075.85420.29410.86590.1600ARI0.79490.74370.846213.58190.67800.81710.4951BMI*WC0.80290.75130.85451932.61210.73680.79270.5295BMI*ABSI0.79040.73930.84141800.36700.65940.82930.4887OBMI+ABSI0.79500.74380.846233.18800.67800.81710.4951OBMI+WC0.80230.75090.853639.74760.71210.81710.5292Abbreviations as in Table 1***P < 0.05 compared with BMI*WC in each age group of both genders\\n\\nThe area under the curve, best threshold, sensitivities, specificities, and Youden’s index of BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, BMI, WC, ABSI, and ARI as a predictor of NAFLD in the different age groups of both genders\\nAbbreviations as in Table 1\\n***P < 0.05 compared with BMI*WC in each age group of both genders\",\n \"Our main findings showed that BMI, WC, ABSI, ARI, BMI*WC, BMI*ABSI, OBMI+ABSI, OBMI+WC, and the obesity phenotypes were all significantly and positively correlated with NAFLD risk; and for the first time in the general population, it was found that BMI combined with WC and ABSI can significantly improve the ability of the basic anthropometric measures to identify NAFLD, especially in middle-aged females and young and middle-aged males. Of significant mention, compared with other age stages in both sexes, all obesity indicators in this study have the highest diagnostic efficiency among young females.\\nIn recent years, the rapid global economic development, dietary patterns and lifestyles of the residents have changed dramatically, and a high-calorie, high-fat daily diet and a sedentary lifestyle have become the main themes of today’s society; therefore, the prevalence of metabolic syndrome, obesity, and other obesity-related diseases has increased greatly [2, 32, 33]. NAFLD is considered the hepatic manifestation of the metabolic syndrome and encompasses a complex spectrum of liver pathologies ranging from simple hepatic steatosis to non-alcoholic steatohepatitis, and also the most common chronic liver disease and the leading cause of liver-related death worldwide [1, 34]. The occurrence and development of NAFLD are closely related to the global epidemic of obesity [6]. Recent epidemiological surveys have shown that the incidence of NAFLD in obese patients can reach 80%, and in severely obese patients who require bariatric surgery, the incidence of NAFLD is as high as 90% [4].\\nIndependent associations between NAFLD and traditional obesity indicators BMI and WC have been demonstrated in many observational studies, but there is much debate as to which indicator is better at identifying obesity-related NAFLD risk [7–10, 35–39]. Several recent studies have demonstrated that BMI was a stronger indicator of obesity for identifying and predicting NAFLD risk, with BMI having the largest AUC compared with other obesity-related parameters, showing better diagnostic performance than WC and other obesity indicators [37–39]. Of course, there were other studies that expressed the opposite standpoints; two studies of NAFLD in obese adolescents and children found that both WC and BMI independently predicted NAFLD risk, with WC, in particular, having the highest predictive accuracy [35, 36]. Furthermore, a meta-analysis study by Qing et al. found that WC remained the strongest anthropometric predictor of NAFLD even after adjusting for a large number of confounding factors, including BMI, and that central obesity was a more dangerous pattern of body fat distribution than general obesity [7]. Clearly, both BMI and WC were independent risk factors for NAFLD, but neither BMI nor WC alone appeared to be an adequate measure of obesity-related NAFLD risk. In a study by Janssen et al. [40], it was found that either BMI or WC alone was an independent predictor of total fat, non-abdominal fat, abdominal subcutaneous fat, and visceral fat, but 20–40% of the variation in fat mass in these fat pools remained unexplained by BMI or WC alone; after they further combining BMI with WC, found that the variability of fat content in these fat pools decreased significantly. Given that the available evidence generally supports both abdominal subcutaneous fat and visceral fat as independent predictors of insulin resistance [41, 42], we hypothesized that the combination of WC and BMI may be a better predictor of metabolic disease. This speculation has been confirmed in studies of metabolic diseases such as cardiovascular disease, obesity-related hypertension, type 2 diabetes, and risk of all-cause mortality [14–17]. Furthermore, in a recent study, Wang et al. also found that BMI combined with WC was more strongly associated with NAFLD in a diabetic population than either BMI or WC alone [18]. However, their study subjects were limited to the diabetic population and did not explore the value of combined BMI and WC in the identification of NAFLD. In addition, it is also worth noting that in the above-mentioned studies related to BMI combined with WC, almost all researchers ignored the statistical artifacts that may result from the strong correlation between BMI and WC. A recent study by Christakoudi S et al. found a traditional U-shaped association between BMI alone and all-cause mortality risk, but when BMI was combined with WC, which had a strong correlation, a predominantly negative association was found between BMI and all-cause mortality, i.e., the higher the BMI the lower the mortality risk [19]. It means that the combination of WC and BMI may not simply supplement BMI, but even changed the pattern of associations between BMI and outcome events, which may be a statistical artifact caused by the strong correlation between WC and BMI. In addition, when they combined BMI with the moderately correlated waist index waist-to-height ratio found a similar but more modest change in the pattern of association between BMI and all-cause mortality. Therefore, combining BMI with a correlated abdominal obesity index may lead to biased and possibly even misleading risk estimates and ineffective risk stratification [19]. As expected, when the researchers further combined BMI with the unrelated abdominal obesity index ABSI and hip circumference, they found that the pattern of association between BMI and all-cause mortality did not change and that more fine-grained risk stratification was achieved for ABSI in each BMI stratum. ABSI is a reliable predictor of diseases such as cardiovascular disease and all-cause mortality and a valid proxy for identifying sarcopenic obesity, independent of BMI by design, and is a useful complement to BMI for risk estimation and risk stratification [22, 43–45].\\nIn this study, we also analyzed the correlations between BMI and WC and ABSI, and in agreement with the results of previous studies [19], there was a strong correlation between BMI and WC and almost no correlation between BMI and ABSI (r < 0.05); in addition, in the association analysis of single indicators with NAFLD risk we found that BMI, WC, and ABSI were all significantly and positively associated with the risk of NAFLD (Table 4). Therefore, the combination of BMI and ABSI may be more appropriate than the combination of BMI and WC. Nevertheless, to fully explore the value of combining BMI and the abdominal obesity index to assess the risk of NAFLD and to circumvent possible statistical artifacts, we still combined BMI with WC and ABSI in 4 separate ways based on previous studies: First, the combination of BMI and WC into four obesity phenotypes, BMIN/WCN, BMIO/WCN, BMIN/WCO, and BMIO/WCO, based on the overweight cut-offs for WC and BMI in Asian populations recommended by the WHO Expert Committee [27, 28]. Second, constructing optimal proportional combinations OBMI+WC and OBMI+ABSI based on the regression coefficients of BMI, WC, and ABSI in multivariate logistic regression models. Third, multiplied BMI with WC and ABSI directly to obtain BMI*WC and BMI*ABSI. Fourth, a predictor ARI (BMI, ABSI) was calculated using the method developed by Krakauer NY et al. [26]. In the analysis of the association between each obesity phenotype and NAFLD we found that after adjusting for a large number of confounders (Model 3), subjects with normal BMI and WC in both sexes had the lowest risk of NAFLD, both the BMIO/WCN phenotype and the BMIN/WCO phenotype increased the risk of NAFLD, and subjects with the BMIO/WCO phenotype had the highest risk of NAFLD (male: OR 4.19, 95%CI 3.54, 4.96; female: OR 9.78, 95%CI 7.35, 13.02); these results were in accord with previous studies [18]. Obviously, comprehensive consideration of the effects of BMI and WC on NAFLD in the general population could explain the obesity-related NAFLD risk to a greater extent, identifying more populations at risk for developing NAFLD. Additionally, all other combined indicators in this study were significantly and positively correlated with NAFLD, with BMI*ABSI having the strongest correlation among males and BMI*WC having the strongest correlation with NAFLD among females. The results also did not change after treating BMI*ABSI as a categorical variable, but it is worth mentioning that the Q4 category has a much higher risk of NAFLD in females than in males compared with the Q1 category in BMI*ABSI (OR: female 63.97 vs. male 26.15). Consistently, in the obesity phenotypes, female subjects with BMIO/WCO phenotype also had a significantly higher risk of NAFLD than males (OR: female 9.78 vs. male 4.19). This suggested that elevated BMI and WC and ABSI may be more important risk factors for NAFLD in females, and females should pay more attention to maintaining normal body weight and body shape to prevent the occurrence of NAFLD.\\nOur current study also compared the ability of five continuous combined indicators, ARI, OBMI+WC, OBMI+ABSI, BMI*ABSI, and BMI*WC, to identify NAFLD with BMI and WC and ABSI alone. According to the results of ROC analysis, ARI and OBMI+ABSI had the strongest NAFLD identification ability in both sexes, and although there were only minor differences between the two and OBMI+WC, BMI*WC and BMI*ABSI, they were all significantly higher than BMI and WC and ABSI alone. From Table 5 we found that the optimal proportional combination indicators (OBMI+WC and OBMI+ABSI) have extremely similar diagnostic efficacy to the multiplicative indicators (BMI*WC and BMI*ABSI) and ARI, which was an interesting result. The optimal proportional combination indicators in this study were calculated after directly obtaining the proportional coefficients based on the ratio of the regression coefficients of BMI and WC or ABSI in model 3, which considered the relative weights of the effects of BMI and WC or ABSI on NAFLD and could fully reflect the effects of BMI combined with WC and ABSI on NAFLD. However, the construction of the calculation formula for the optimal proportional combination was relatively complicated, which may be affected by different adjusted confounding factors and changes in the correlation between BMI, WC, ABSI and NAFLD in different test groups. In contrast, the ARI (BMI, ABSI) was obtained by multiplying the logarithms of ORs of BMI and ABSI by the corresponding BMI and ABSI values of each subject and then summing them together [26] and the multiplicative combination indicators were obtained by directly multiplying BMI with WC and ABSI, which are easy to calculate and had a high diagnostic performance in both sexes. In summary, ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be better obesity indicators for clinical screening of NAFLD. Additionally, the results of the ROC analysis after further grouping the sexes by different age groups showed that all combination indicators of BMI and ABSI or WC significantly improved the ability of a single indicator to identify NAFLD in middle-aged females and young and middle-aged males. It is worth mentioning that the highest diagnostic efficacy for NAFLD was found in the young female population for both combined and single obesity indicators; this may be related to the protective effect of large amounts of estrogen on NAFLD in young females. As we all know, the reproductive status of females is closely related to their health status. Experimental studies have shown that estrogens can exert antioxidant, anti-steatosis, and anti-fibrotic effects in the liver [46, 47]. Whereas, a significant decrease in estrogen levels and an increase in circulating androgen levels in postmenopausal females may lead to disorders of lipid metabolism and the development of metabolic syndrome [48]. In addition, the decline in the body’s metabolic status with aging is also an important risk factor for NAFLD [49]. Thus, younger females have a healthier metabolic profile and fewer confounding factors that affect the risk of NAFLD identified by obesity indicators compared with postmenopausal females and males.\\nOur findings have important implications for both clinical screening and epidemiological studies of NAFLD. Due to the insidious onset of NAFLD, early-stage patients may be asymptomatic, primary care workers lack attention to NAFLD, and public awareness and treatment rates are very low [50, 51]; moreover, there are no approved treatments for NAFLD, hence the mainstay of prevention and treatment of NAFLD remains healthy lifestyle and weight control [1]. In the current study, ARI and BMI*WC and BMI*ABSI had greater than 80% diagnostic accuracy in all age groups of both sexes and up to 95.22% in young women, and ARI and BMI*WC and BMI*ABSI were easily calculated without additional measurements or equipment requirements. Therefore, we recommend that the ARI and BMI*ABSI and BMI*WC should be incorporated into the clinical screening plan for NAFLD, which will greatly improve the detection rate of NAFLD.\\n[SUBTITLE] Advantages and limitations of research [SUBSECTION] Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.\\nSeveral advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.\",\n \"Several advantages of the current study need to be mentioned: (1) The current study demonstrated for the first time in a large sample of the general population that BMI combined with WC and ABSI had a stronger ability to identify NAFLD than a basic anthropometric obesity parameter. (2) After combining BMI with WC and ABSI in various ways, we found that the ARI and multiplicative combination indicators (BMI*WC and BMI*ABSI) may be the more appropriate obesity index for clinical screening of NAFLD. These findings provided an important reference for the application of BMI combined with the abdominal obesity index for NAFLD screening in the general population.\\nOf course, it is undeniable that the current study has several limitations: (1) The current study was a cross-sectional study, so we can only demonstrate that BMI combined with the abdominal obesity index was stronger than a single indicator for identifying NAFLD in the general population, but whether the predictive performance for NAFLD risk is stronger needs to be validated in a longitudinal cohort study. (2) The diagnosis of NAFLD, the main outcome in this study, was based on ultrasonography, which may have missed some patients with NAFLD [1]. However, in the setting of fewer NAFLD cases, we still demonstrated that BMI combined with the abdominal obesity index had a higher diagnostic performance for NAFLD than a single indicator. (3) Although we systematically adjusted for known NAFLD risk factors, there may be residual confounding due to the limitations of retrospective studies where some NAFLD risk factors could not be measured and obtained.\",\n \"Taken together, our findings confirmed that BMI combined with ABSI and WC can better explain obesity-related NAFLD risk than a single indicator in the general population, and identified the ARI and the multiplicative combination indicators BMI*ABSI and BMI*WC as the simplest and very effective combination indicators for diagnosing the risk of NAFLD. These new findings underscored the importance of the combined application of BMI and the abdominal obesity index in clinical practice to assess NAFLD risk in the general population and provided a cost-effective tool for precise preventive screening and treatment monitoring for NAFLD.\",\n \"[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\n\\nSupplementary Material 2\\n\\nSupplementary Material 2\\nBelow is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\n\\nSupplementary Material 2\\n\\nSupplementary Material 2\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\n\\nSupplementary Material 2\\n\\nSupplementary Material 2\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,"results",null,null,null,null,"discussion",null,"conclusion","supplementary-material",null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n \"conclusion\",\n \"supplementary-material\",\n null\n]"},"keywords":{"kind":"list like","value":["Body mass index","Waist circumference","ABSI","ARI","Receiver operating characteristic analysis","Non-alcoholic fatty liver disease"],"string":"[\n \"Body mass index\",\n \"Waist circumference\",\n \"ABSI\",\n \"ARI\",\n \"Receiver operating characteristic analysis\",\n \"Non-alcoholic fatty liver disease\"\n]"}}},{"rowIdx":374,"cells":{"title":{"kind":"string","value":"The blacksmith approach: a strategy for teaching and learning in the medical anatomy course (a qualitative study)."},"pmid":{"kind":"string","value":"36266705"},"background_abstract":{"kind":"string","value":"Anatomy is a symbolic, essential core topic and one of the fundamental pillars of medical and paramedical knowledge. Nevertheless, few exploratory data analyses have focused on how students approach learning anatomy. This study examined how students perceive their learning experience during anatomy lessons and how to make a model which promotes their meaningful learning and professional identity."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Using purposive sampling with maximum variation, we conducted a qualitative content analysis at the Shiraz University of Medical Sciences in Iran (2020 to 2021). Twenty-four medical students and twelve faculty members of Iran's medical science universities were enrolled in the study. The data were collected through semi-structured interviews and analyzed according to the theme."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"A conceptual model emerged from the data analysis with the main theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (adequate preparation and mindful beginning), (2) heating the students' hearts (considering supporting systems that learners need) and (3) using Sledgehammer's approach (teaching anatomy by using more active methods and engaging all neuroanatomical regions) and (Using fun for enjoyable learning). All the concepts were related to each other."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Medical students experience a challenging fundamental evolution into professional doctors. Educational systems focus primarily on teaching and learning, while students' transition can be facilitated by a three-step model called the Blacksmith Approach. It best serves as an educational framework for any pivotal, preclinical course capable of helping students acquire new roles and tackle challenges. Further research should be conducted to confirm how hard work leads to satisfying results with the opportunity to create enjoyable learning."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Learning","Students, Medical","Education, Medical, Undergraduate","Qualitative Research","Curriculum","Teaching","Anatomy"],"string":"[\n \"Humans\",\n \"Learning\",\n \"Students, Medical\",\n \"Education, Medical, Undergraduate\",\n \"Qualitative Research\",\n \"Curriculum\",\n \"Teaching\",\n \"Anatomy\"\n]"},"pmcid":{"kind":"string","value":"9584281"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Anatomy, as a branch of biological and medical sciences, is one of the introductory courses in the learning experience of undergraduate medical students, which is essential for impending clinical practice and clinical proficiency [1, 2]. A challenging question facing medical education is how to change some young students into responsible and resilient professional physicians. Most of the answers are being placed in clinical years; however, we can begin the transition process earlier with a promising tool called anatomy [3, 4].\nDespite being one of the fundamental pillars of medical knowledge, anatomy has been pulled back as a simple course in the preclinical period. Historically, ancient medical schools have relied heavily on philosophical and intuitive systems of thought [5]. Experimental investigations by dissection helped physicians actualize the human body as a combination of multiple organ systems. [6]. We can assume that anatomy science has led to reform in how medicine views the human being from a spiritual, holistic being into a biopsychosocial person. However, anatomy has been underestimated in current medical pedagogy [7, 8].\nRecent evidence indicates that we can reform anatomy courses making them as influential as they can be. [9–11]. Anatomy and dissection sessions can act as students’ first encounters with the humanized face of medicine, where medical students can learn teamwork, morality, coping strategies, and communication skills. In addition, anatomy can be perceived as a way of internalizing self-awareness, medical epistemology, empathy, and medical ethics [6, 12, 13].\nWhile a myriad of teaching approaches are available for anatomy education, they have changed according to the needs of learners or global crises like covid-19. The primary approaches include lecture, dissection, and demonstration of cadavers. In addition, many schools incorporated radiological imaging as a learning tool for teaching anatomy in the living body. Considering the shortcomings of traditional methods, medical schools started implementing more innovative techniques that enhanced active learning. While Problem-based learning, flipped classroom, and reflective writing activities on the cadaveric dissection fostered communication skills and self-directed learning [12, 14, 15], other methods focused on students’ learning styles: drawing and whiteboarding, three-dimensional printing (3DP), and digital models for visual learners; play-doh, role modeling and peer physical exam for tactile learning; singing, dancing, yoga, and pilates for kinaesthetic learners [16–18]. The covid pandemic disrupted anatomy education as the number of donated cadavers decreased substantially, and social distancing prevented in-person classes or labs [17–19]. Accordingly, the teaching approaches relied, more than ever, on technology leading to the widespread use of computer-assisted learning, web-based learning via social media (esp. Facebook and youtube), and a more prominent role for Augmented-Reality and Virtual-Reality [20, 21]. The main focus behind all these methods and strategies is better teaching anatomy for its sake. However, such an approach does not allow us to use the full potential of anatomy science as a core element in changing young students into self-confident and resilient learners, which leads to the creation of future professional physicians [22, 23].\nPhysicians need professional competencies, a set of which essential competencies, including professional resilience, is and emotional competencies, to achieve goals. As a capacity to endure difficulties and quickly improve from a stress-inducing experience, resilience helps one to make valuable adaptations to problems and can utilize as a protective defense against adversities, thus facilitating welfare [23, 24]. Resilience is referred to as one’s capability to maintain and improve their well-being when faced with life challenges. Resilience can be considered a behavior acquired during training and various courses, including the main course of anatomy. Resilience includes cognitive processes and has four dimensions, including self-efficacy. Planning, Self-control, Commitment, and perseverance include the following: young doctors must be present in various educational environments and support clinical and educational supervisors in enduring difficulties to help them develop personal and professional resilience [24, 25].\nIn Iran and at Shiraz University, anatomy courses entail three steps. Initially, a topic is introduced through face-to-face, online or offline lectures with a slide or a short video clip. Next, students work on anatomical modeling and, finally, on a cadaver [26, 27].\nToday, not only the medical students must bear a significant load of details in anatomy courses, which can reduce their motivation, but they should also face the educational consequences of the covid pandemic, which include loss of hands-on learning activities and limited access to tools like cadavers, models, pathology specimens, and skeletons [28–34]. Losing face-to-face contact and direct interaction with peers and teachers may hinder the students’ growth as future physicians [6, 13]. Therefore, finding ways to improve the learning experience is more challenging than ever."},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Participants’ characteristics [SUBSECTION] In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem\nIn this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem\n[SUBTITLE] Central theme: blacksmith approach [SUBSECTION] According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\nAccording to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n[SUBTITLE] Sub theme1: redesigning educational mindset (making a new forge) [SUBSECTION] Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\nCreating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n[SUBTITLE] Subtheme 2: role modeling and student support (heating up the students’ hearts) [SUBSECTION] The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\nThe professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n[SUBTITLE] Joyful teaching (telling stories/games/humor) [SUBSECTION] According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\nAccording to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n[SUBTITLE] Student Support System [SUBSECTION] Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\nBoth faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n[SUBTITLE] Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics) [SUBSECTION] While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\nWhile anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\n[SUBTITLE] Study guide [SUBSECTION] Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\nMost participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n[SUBTITLE] Neuroanatomical engagement [SUBSECTION] Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\nParticipants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n[SUBTITLE] Reference autonomy [SUBSECTION] According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\nAccording to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n[SUBTITLE] Mindful dissection, better transition [SUBSECTION] The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\nThe first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n[SUBTITLE] Drawing [SUBSECTION] Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\nTeachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n[SUBTITLE] Problem-based learning [SUBSECTION] Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\nStudents will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n[SUBTITLE] Peer-assisted learning [SUBSECTION] Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\nStudents stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"Medical students experience a fundamental evolution into responsible and lifelong learners qualified as professional doctors. Educational systems focus primarily on teaching and learning, while students’ transition can be facilitated by a three-step model called the Blacksmith Approach, based on the study results. Initially, the blacksmith (educational system) must prepare a new forge. It means anatomy teachers should redesign the academic mentality of students by delivering messages about role change, difficulties ahead, and the importance of self-guidance, along with implementing clinical scenarios in active learning methods. Then, he/she must heat the metal. It implies that anatomy teachers, as gatekeepers of medicine, should adopt amusing teaching methods while embracing their role modeling and actively applying support strategies. Finally, he/she may shape the metal with a sledgehammer. Teachers need to use three components to create a sword (self-confident, accountable, resilient professional); thorough engagement by using active and engaging teaching techniques, reframing the situation with problem-based clinical scenarios, and joyful teaching. And selective information seeking via study guide and reference autonomy.\nIn conclusion, an active and integrated strategy for teaching and learning anatomy is suggested to be applied in other courses and areas of medical education in the current medical curriculum."},"other_sections_titles":{"kind":"list like","value":["Objectives","Design","Setting","Sampling","Data collection","Data analysis","Participants’ characteristics","Central theme: blacksmith approach","Sub theme1: redesigning educational mindset (making a new forge)","Subtheme 2: role modeling and student support (heating up the students’ hearts)","Joyful teaching (telling stories/games/humor)","Student Support System","Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics)","Study guide","Neuroanatomical engagement","Reference autonomy","Mindful dissection, better transition","Drawing","Problem-based learning","Peer-assisted learning","First Step: A new forge (mindset)","Second step","Third step: Sledgehammer’s approach","Study limitation",""],"string":"[\n \"Objectives\",\n \"Design\",\n \"Setting\",\n \"Sampling\",\n \"Data collection\",\n \"Data analysis\",\n \"Participants’ characteristics\",\n \"Central theme: blacksmith approach\",\n \"Sub theme1: redesigning educational mindset (making a new forge)\",\n \"Subtheme 2: role modeling and student support (heating up the students’ hearts)\",\n \"Joyful teaching (telling stories/games/humor)\",\n \"Student Support System\",\n \"Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics)\",\n \"Study guide\",\n \"Neuroanatomical engagement\",\n \"Reference autonomy\",\n \"Mindful dissection, better transition\",\n \"Drawing\",\n \"Problem-based learning\",\n \"Peer-assisted learning\",\n \"First Step: A new forge (mindset)\",\n \"Second step\",\n \"Third step: Sledgehammer’s approach\",\n \"Study limitation\",\n \"\"\n]"},"other_sections_texts":{"kind":"list like","value":["In this study, we aimed to identify the themes that explain the strategy for teaching and learning and foster meaningful learning and professional identity in medical anatomy courses among medical students in Iran. Also, we aimed to develop a conceptual framework to explain this strategy development based on the experience of both students and faculty members(FM) through a qualitative study.","A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.","This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].","We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.","We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\n\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\n\nWhat is your description of successful teaching/learning in an anatomy course?\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.","We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].","In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem","According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).","Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).","The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)","According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)","Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)","While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.","Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)","Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)","According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)","The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)","Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)","Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)","Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)","must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].","The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].","We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.","As this study was conducted in one country, it is suggested that more studies and insights be obtained from other participants to increase the generalizability of findings. Besides, further work should be done to provide more practical guidance in the educational system regarding the transition of medical students in teaching and learning anatomy in the current pandemic crisis.","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"In this study, we aimed to identify the themes that explain the strategy for teaching and learning and foster meaningful learning and professional identity in medical anatomy courses among medical students in Iran. Also, we aimed to develop a conceptual framework to explain this strategy development based on the experience of both students and faculty members(FM) through a qualitative study.\",\n \"A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\",\n \"This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\",\n \"We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\",\n \"We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\\n\\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\n\\nWhat is your description of successful teaching/learning in an anatomy course?\\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\",\n \"We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].\",\n \"In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\\n\\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\\nstudents\\n\\nGender\\nNoPercent (%)\\nMean Age (y ± SD)\\nMale1562.522.3 ± 8.4Female937.5\\nDegree\\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\\n\\nDemographic Characteristics of the Study Participants\\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\\n\\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\\nBlacksmith Approach\\n\\nRedesigning Educational Mindset (Making a New Forge)\\n\\nCreating a Vision for Students\\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\\nMindful Clinical View\\n\\nRole Modeling and Support (Heating up the Students’ Hearts)\\n\\nJoyful Teaching (Telling Stories/Games/ Humor)\\n√ every trick must be done√ enjoy class√ follow me in the classroom\\nStudent Support System\\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\\n\\nStudy Guide\\n√ Preparedness√ Road map\\nReference Autonomy\\n√ self-learning√ self-direction\\nNeuroanatomical Engagement\\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\\nMindful Dissection, Better Transition\\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\\nDrawing\\n√ easy to learn√ Draw a lot\\nProblem-Based Learning\\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\\nPeer-Assisted Learning\\n√ Peer to the peer group√ Discussing the problem\\n\\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\\n√ Anatomy is hard, bulky, should be taken seriously\\n√ I can help myself more than anyone\\n√ integrated curriculum\\n√ self-study / Self-confidence\\nscarcely happens in the real world!\\n√ professional identity\\n√ prospective physicians\\n√ every trick must be done\\n√ enjoy class\\n√ follow me in the classroom\\n√ susceptible students\\n√ assistance\\n√ Vulnerable\\n√ right supporters\\nSitting and talking to them\\n√ Preparedness\\n√ Road map\\n√ self-learning\\n√ self-direction\\n√ Visual Learning\\n√ Auditory Learning\\n√ Sensory Learning\\n√ Brain Hemispheres\\nLong-term Memorization\\n√ It smells so bad, but we need the observation\\n√ We got used to it\\n√ ethically important\\n√ respect the dead body\\n√ respect to the living body\\n√ easy to learn\\n√ Draw a lot\\n√ Look at the patient’s family\\n√ Deep worry\\n√ What is the next step of the problem?\\n√ Peer to the peer group\\n√ Discussing the problem\",\n \"According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\\n\\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\n\\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\",\n \"Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\\n\\nCreating a Vision for Students.\\n\\nCreating a Vision for Students.\\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n\\n2)Mindful Clinical View.\\n\\nMindful Clinical View.\\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\",\n \"The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\",\n \"According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\",\n \"Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\",\n \"While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\",\n \"Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\",\n \"Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\",\n \"According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\",\n \"The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\",\n \"Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\",\n \"Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\",\n \"Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\",\n \"must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\",\n \"The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\",\n \"We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.\",\n \"As this study was conducted in one country, it is suggested that more studies and insights be obtained from other participants to increase the generalizability of findings. Besides, further work should be done to provide more practical guidance in the educational system regarding the transition of medical students in teaching and learning anatomy in the current pandemic crisis.\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Objectives","Materials and methods","Design","Setting","Sampling","Data collection","Data analysis","Results","Participants’ characteristics","Central theme: blacksmith approach","Sub theme1: redesigning educational mindset (making a new forge)","Subtheme 2: role modeling and student support (heating up the students’ hearts)","Joyful teaching (telling stories/games/humor)","Student Support System","Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics)","Study guide","Neuroanatomical engagement","Reference autonomy","Mindful dissection, better transition","Drawing","Problem-based learning","Peer-assisted learning","Discussion","First Step: A new forge (mindset)","Second step","Third step: Sledgehammer’s approach","Study limitation","Conclusion","Electronic supplementary material",""],"string":"[\n \"Introduction\",\n \"Objectives\",\n \"Materials and methods\",\n \"Design\",\n \"Setting\",\n \"Sampling\",\n \"Data collection\",\n \"Data analysis\",\n \"Results\",\n \"Participants’ characteristics\",\n \"Central theme: blacksmith approach\",\n \"Sub theme1: redesigning educational mindset (making a new forge)\",\n \"Subtheme 2: role modeling and student support (heating up the students’ hearts)\",\n \"Joyful teaching (telling stories/games/humor)\",\n \"Student Support System\",\n \"Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics)\",\n \"Study guide\",\n \"Neuroanatomical engagement\",\n \"Reference autonomy\",\n \"Mindful dissection, better transition\",\n \"Drawing\",\n \"Problem-based learning\",\n \"Peer-assisted learning\",\n \"Discussion\",\n \"First Step: A new forge (mindset)\",\n \"Second step\",\n \"Third step: Sledgehammer’s approach\",\n \"Study limitation\",\n \"Conclusion\",\n \"Electronic supplementary material\",\n \"\"\n]"},"all_sections_texts":{"kind":"list like","value":["Anatomy, as a branch of biological and medical sciences, is one of the introductory courses in the learning experience of undergraduate medical students, which is essential for impending clinical practice and clinical proficiency [1, 2]. A challenging question facing medical education is how to change some young students into responsible and resilient professional physicians. Most of the answers are being placed in clinical years; however, we can begin the transition process earlier with a promising tool called anatomy [3, 4].\nDespite being one of the fundamental pillars of medical knowledge, anatomy has been pulled back as a simple course in the preclinical period. Historically, ancient medical schools have relied heavily on philosophical and intuitive systems of thought [5]. Experimental investigations by dissection helped physicians actualize the human body as a combination of multiple organ systems. [6]. We can assume that anatomy science has led to reform in how medicine views the human being from a spiritual, holistic being into a biopsychosocial person. However, anatomy has been underestimated in current medical pedagogy [7, 8].\nRecent evidence indicates that we can reform anatomy courses making them as influential as they can be. [9–11]. Anatomy and dissection sessions can act as students’ first encounters with the humanized face of medicine, where medical students can learn teamwork, morality, coping strategies, and communication skills. In addition, anatomy can be perceived as a way of internalizing self-awareness, medical epistemology, empathy, and medical ethics [6, 12, 13].\nWhile a myriad of teaching approaches are available for anatomy education, they have changed according to the needs of learners or global crises like covid-19. The primary approaches include lecture, dissection, and demonstration of cadavers. In addition, many schools incorporated radiological imaging as a learning tool for teaching anatomy in the living body. Considering the shortcomings of traditional methods, medical schools started implementing more innovative techniques that enhanced active learning. While Problem-based learning, flipped classroom, and reflective writing activities on the cadaveric dissection fostered communication skills and self-directed learning [12, 14, 15], other methods focused on students’ learning styles: drawing and whiteboarding, three-dimensional printing (3DP), and digital models for visual learners; play-doh, role modeling and peer physical exam for tactile learning; singing, dancing, yoga, and pilates for kinaesthetic learners [16–18]. The covid pandemic disrupted anatomy education as the number of donated cadavers decreased substantially, and social distancing prevented in-person classes or labs [17–19]. Accordingly, the teaching approaches relied, more than ever, on technology leading to the widespread use of computer-assisted learning, web-based learning via social media (esp. Facebook and youtube), and a more prominent role for Augmented-Reality and Virtual-Reality [20, 21]. The main focus behind all these methods and strategies is better teaching anatomy for its sake. However, such an approach does not allow us to use the full potential of anatomy science as a core element in changing young students into self-confident and resilient learners, which leads to the creation of future professional physicians [22, 23].\nPhysicians need professional competencies, a set of which essential competencies, including professional resilience, is and emotional competencies, to achieve goals. As a capacity to endure difficulties and quickly improve from a stress-inducing experience, resilience helps one to make valuable adaptations to problems and can utilize as a protective defense against adversities, thus facilitating welfare [23, 24]. Resilience is referred to as one’s capability to maintain and improve their well-being when faced with life challenges. Resilience can be considered a behavior acquired during training and various courses, including the main course of anatomy. Resilience includes cognitive processes and has four dimensions, including self-efficacy. Planning, Self-control, Commitment, and perseverance include the following: young doctors must be present in various educational environments and support clinical and educational supervisors in enduring difficulties to help them develop personal and professional resilience [24, 25].\nIn Iran and at Shiraz University, anatomy courses entail three steps. Initially, a topic is introduced through face-to-face, online or offline lectures with a slide or a short video clip. Next, students work on anatomical modeling and, finally, on a cadaver [26, 27].\nToday, not only the medical students must bear a significant load of details in anatomy courses, which can reduce their motivation, but they should also face the educational consequences of the covid pandemic, which include loss of hands-on learning activities and limited access to tools like cadavers, models, pathology specimens, and skeletons [28–34]. Losing face-to-face contact and direct interaction with peers and teachers may hinder the students’ growth as future physicians [6, 13]. Therefore, finding ways to improve the learning experience is more challenging than ever.","In this study, we aimed to identify the themes that explain the strategy for teaching and learning and foster meaningful learning and professional identity in medical anatomy courses among medical students in Iran. Also, we aimed to develop a conceptual framework to explain this strategy development based on the experience of both students and faculty members(FM) through a qualitative study.","[SUBTITLE] Design [SUBSECTION] A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\nA qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\n[SUBTITLE] Setting [SUBSECTION] This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\nThis study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\n[SUBTITLE] Sampling [SUBSECTION] We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\nWe used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\n[SUBTITLE] Data collection [SUBSECTION] We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\n\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\n\nWhat is your description of successful teaching/learning in an anatomy course?\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\nWe utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\n\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\n\nWhat is your description of successful teaching/learning in an anatomy course?\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\n[SUBTITLE] Data analysis [SUBSECTION] We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].\nWe listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].","A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.","This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].","We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.","We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\n\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\n\nWhat is your description of successful teaching/learning in an anatomy course?\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.","We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].","[SUBTITLE] Participants’ characteristics [SUBSECTION] In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem\nIn this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem\n[SUBTITLE] Central theme: blacksmith approach [SUBSECTION] According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\nAccording to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n[SUBTITLE] Sub theme1: redesigning educational mindset (making a new forge) [SUBSECTION] Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\nCreating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n[SUBTITLE] Subtheme 2: role modeling and student support (heating up the students’ hearts) [SUBSECTION] The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\nThe professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n[SUBTITLE] Joyful teaching (telling stories/games/humor) [SUBSECTION] According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\nAccording to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n[SUBTITLE] Student Support System [SUBSECTION] Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\nBoth faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n[SUBTITLE] Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics) [SUBSECTION] While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\nWhile anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\n[SUBTITLE] Study guide [SUBSECTION] Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\nMost participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n[SUBTITLE] Neuroanatomical engagement [SUBSECTION] Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\nParticipants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n[SUBTITLE] Reference autonomy [SUBSECTION] According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\nAccording to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n[SUBTITLE] Mindful dissection, better transition [SUBSECTION] The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\nThe first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n[SUBTITLE] Drawing [SUBSECTION] Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\nTeachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n[SUBTITLE] Problem-based learning [SUBSECTION] Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\nStudents will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n[SUBTITLE] Peer-assisted learning [SUBSECTION] Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\nStudents stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)","In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\n\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\nstudents\n\nGender\nNoPercent (%)\nMean Age (y ± SD)\nMale1562.522.3 ± 8.4Female937.5\nDegree\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\n\nDemographic Characteristics of the Study Participants\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\n\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\nBlacksmith Approach\n\nRedesigning Educational Mindset (Making a New Forge)\n\nCreating a Vision for Students\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\nMindful Clinical View\n\nRole Modeling and Support (Heating up the Students’ Hearts)\n\nJoyful Teaching (Telling Stories/Games/ Humor)\n√ every trick must be done√ enjoy class√ follow me in the classroom\nStudent Support System\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\n\nStudy Guide\n√ Preparedness√ Road map\nReference Autonomy\n√ self-learning√ self-direction\nNeuroanatomical Engagement\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\nMindful Dissection, Better Transition\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\nDrawing\n√ easy to learn√ Draw a lot\nProblem-Based Learning\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\nPeer-Assisted Learning\n√ Peer to the peer group√ Discussing the problem\n\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\n√ Anatomy is hard, bulky, should be taken seriously\n√ I can help myself more than anyone\n√ integrated curriculum\n√ self-study / Self-confidence\nscarcely happens in the real world!\n√ professional identity\n√ prospective physicians\n√ every trick must be done\n√ enjoy class\n√ follow me in the classroom\n√ susceptible students\n√ assistance\n√ Vulnerable\n√ right supporters\nSitting and talking to them\n√ Preparedness\n√ Road map\n√ self-learning\n√ self-direction\n√ Visual Learning\n√ Auditory Learning\n√ Sensory Learning\n√ Brain Hemispheres\nLong-term Memorization\n√ It smells so bad, but we need the observation\n√ We got used to it\n√ ethically important\n√ respect the dead body\n√ respect to the living body\n√ easy to learn\n√ Draw a lot\n√ Look at the patient’s family\n√ Deep worry\n√ What is the next step of the problem?\n√ Peer to the peer group\n√ Discussing the problem","According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\n\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\n\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).","Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\n\nCreating a Vision for Students.\n\nCreating a Vision for Students.\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \" (Student no. 1).\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\n\n2)Mindful Clinical View.\n\nMindful Clinical View.\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).","The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)","According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)","Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)","While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.","Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)","Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)","According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)","The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)","Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)","Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \" (FM no. 6)","Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)","We first scrutinized the experience of teaching and learning anatomy through qualitative analysis. Next, we conceptualized a new approach to help medical students in anatomy courses have more satisfactory learning outcomes and a better transition into resilient future doctors. We called it “the Blacksmith Approach.“ building a sword (prosperous and resilient learner) takes three steps. 1. Creating a new forge (adequate preparation and mindful beginning) 2; heating students’ hearts (support and role modeling) 3. Sledgehammer’s approach (thorough engagement in the age of pandemics). (Fig. 1)\nWhile other proposed educational methods in teaching anatomy focus mainly on learning outcomes during anatomy courses [14–19], the blacksmith approach tries to reach two goals, i.e., better teaching-learning outcomes and helping young students transition into future medical professionals. In fact, this is not another teaching method. It best serves as an educational framework for any pivotal, preclinical course capable of helping students acquire new roles and tackle challenges. In this sense, medical educators can adopt all prior teaching techniques as long as they help students reach better learning outcomes or have better transition experiences.\nThe blacksmith approach can be an educational implication of Schlossberg’s 4 S model for transition. Transition is a process that enables individuals to incorporate change into their life. The 4 S theory argues that an individual goes through a cyclic experience of 4 stages during the transition, i.e., situation, self, support, and strategies [46]. We encounter two types of transition in an educational setting. One happens within the learners and one within the educators. Although more attention is put on the former, we cannot achieve desired outcomes if we neglect the latter. While in the 4 S model, these four stages happen within the individual, the Blacksmith Approach creates a compatible, three-step learning environment for students and teachers to transition by creating a new attitude and mindset for both groups. As Browne et al. stated, individual motivation, background and circumstances, a sense of control, organizational support, effective networking, and information-seeking behavior contributed to a successful transition into and maintaining a strong self-identity as a medical educator [47].\n[SUBTITLE] First Step: A new forge (mindset) [SUBSECTION] must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\nmust be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\n[SUBTITLE] Second step [SUBSECTION] The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\nThe students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\n[SUBTITLE] Third step: Sledgehammer’s approach [SUBSECTION] We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.\nWe found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.","must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].","The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].","We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.","As this study was conducted in one country, it is suggested that more studies and insights be obtained from other participants to increase the generalizability of findings. Besides, further work should be done to provide more practical guidance in the educational system regarding the transition of medical students in teaching and learning anatomy in the current pandemic crisis.","Medical students experience a fundamental evolution into responsible and lifelong learners qualified as professional doctors. Educational systems focus primarily on teaching and learning, while students’ transition can be facilitated by a three-step model called the Blacksmith Approach, based on the study results. Initially, the blacksmith (educational system) must prepare a new forge. It means anatomy teachers should redesign the academic mentality of students by delivering messages about role change, difficulties ahead, and the importance of self-guidance, along with implementing clinical scenarios in active learning methods. Then, he/she must heat the metal. It implies that anatomy teachers, as gatekeepers of medicine, should adopt amusing teaching methods while embracing their role modeling and actively applying support strategies. Finally, he/she may shape the metal with a sledgehammer. Teachers need to use three components to create a sword (self-confident, accountable, resilient professional); thorough engagement by using active and engaging teaching techniques, reframing the situation with problem-based clinical scenarios, and joyful teaching. And selective information seeking via study guide and reference autonomy.\nIn conclusion, an active and integrated strategy for teaching and learning anatomy is suggested to be applied in other courses and areas of medical education in the current medical curriculum.","[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\nBelow is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"Anatomy, as a branch of biological and medical sciences, is one of the introductory courses in the learning experience of undergraduate medical students, which is essential for impending clinical practice and clinical proficiency [1, 2]. A challenging question facing medical education is how to change some young students into responsible and resilient professional physicians. Most of the answers are being placed in clinical years; however, we can begin the transition process earlier with a promising tool called anatomy [3, 4].\\nDespite being one of the fundamental pillars of medical knowledge, anatomy has been pulled back as a simple course in the preclinical period. Historically, ancient medical schools have relied heavily on philosophical and intuitive systems of thought [5]. Experimental investigations by dissection helped physicians actualize the human body as a combination of multiple organ systems. [6]. We can assume that anatomy science has led to reform in how medicine views the human being from a spiritual, holistic being into a biopsychosocial person. However, anatomy has been underestimated in current medical pedagogy [7, 8].\\nRecent evidence indicates that we can reform anatomy courses making them as influential as they can be. [9–11]. Anatomy and dissection sessions can act as students’ first encounters with the humanized face of medicine, where medical students can learn teamwork, morality, coping strategies, and communication skills. In addition, anatomy can be perceived as a way of internalizing self-awareness, medical epistemology, empathy, and medical ethics [6, 12, 13].\\nWhile a myriad of teaching approaches are available for anatomy education, they have changed according to the needs of learners or global crises like covid-19. The primary approaches include lecture, dissection, and demonstration of cadavers. In addition, many schools incorporated radiological imaging as a learning tool for teaching anatomy in the living body. Considering the shortcomings of traditional methods, medical schools started implementing more innovative techniques that enhanced active learning. While Problem-based learning, flipped classroom, and reflective writing activities on the cadaveric dissection fostered communication skills and self-directed learning [12, 14, 15], other methods focused on students’ learning styles: drawing and whiteboarding, three-dimensional printing (3DP), and digital models for visual learners; play-doh, role modeling and peer physical exam for tactile learning; singing, dancing, yoga, and pilates for kinaesthetic learners [16–18]. The covid pandemic disrupted anatomy education as the number of donated cadavers decreased substantially, and social distancing prevented in-person classes or labs [17–19]. Accordingly, the teaching approaches relied, more than ever, on technology leading to the widespread use of computer-assisted learning, web-based learning via social media (esp. Facebook and youtube), and a more prominent role for Augmented-Reality and Virtual-Reality [20, 21]. The main focus behind all these methods and strategies is better teaching anatomy for its sake. However, such an approach does not allow us to use the full potential of anatomy science as a core element in changing young students into self-confident and resilient learners, which leads to the creation of future professional physicians [22, 23].\\nPhysicians need professional competencies, a set of which essential competencies, including professional resilience, is and emotional competencies, to achieve goals. As a capacity to endure difficulties and quickly improve from a stress-inducing experience, resilience helps one to make valuable adaptations to problems and can utilize as a protective defense against adversities, thus facilitating welfare [23, 24]. Resilience is referred to as one’s capability to maintain and improve their well-being when faced with life challenges. Resilience can be considered a behavior acquired during training and various courses, including the main course of anatomy. Resilience includes cognitive processes and has four dimensions, including self-efficacy. Planning, Self-control, Commitment, and perseverance include the following: young doctors must be present in various educational environments and support clinical and educational supervisors in enduring difficulties to help them develop personal and professional resilience [24, 25].\\nIn Iran and at Shiraz University, anatomy courses entail three steps. Initially, a topic is introduced through face-to-face, online or offline lectures with a slide or a short video clip. Next, students work on anatomical modeling and, finally, on a cadaver [26, 27].\\nToday, not only the medical students must bear a significant load of details in anatomy courses, which can reduce their motivation, but they should also face the educational consequences of the covid pandemic, which include loss of hands-on learning activities and limited access to tools like cadavers, models, pathology specimens, and skeletons [28–34]. Losing face-to-face contact and direct interaction with peers and teachers may hinder the students’ growth as future physicians [6, 13]. Therefore, finding ways to improve the learning experience is more challenging than ever.\",\n \"In this study, we aimed to identify the themes that explain the strategy for teaching and learning and foster meaningful learning and professional identity in medical anatomy courses among medical students in Iran. Also, we aimed to develop a conceptual framework to explain this strategy development based on the experience of both students and faculty members(FM) through a qualitative study.\",\n \"[SUBTITLE] Design [SUBSECTION] A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\\nA qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\\n[SUBTITLE] Setting [SUBSECTION] This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\\nThis study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\\n[SUBTITLE] Sampling [SUBSECTION] We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\\nWe used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\\n[SUBTITLE] Data collection [SUBSECTION] We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\\n\\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\n\\nWhat is your description of successful teaching/learning in an anatomy course?\\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\\nWe utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\\n\\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\n\\nWhat is your description of successful teaching/learning in an anatomy course?\\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\\n[SUBTITLE] Data analysis [SUBSECTION] We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].\\nWe listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].\",\n \"A qualitative approach was used by inductive content analysis in the first step [35]. Then we formulated the discovered themes into a comprehensive pedagogical approach.\",\n \"This study was conducted at Shiraz University of Medical Sciences (SUMS). The university currently includes more than 10,000 students, 200 majors, 782 faculty members, 54 research centers, 13 educational hospitals, a history of 70 years, and an assertive group teaching anatomy with fourteen faculty members in different academic ranks (assistant professors, associated professors, and full professors).\\nMedical students start an anatomy course as they enter medical school with an integrated organ-based curriculum. They pass 15 academic and practical units of this course, including complete systems: blood circulation, respiration, endocrine, musculoskeletal, and urinary, in integration with other classes such as physiology, biochemistry, and clinical courses on the same topics. Students participate in formative and summative assessment tests such as quizzes and final examinations to improve learning and obtain end-of-semester grades. Students should reveal in-depth, thoughtful principles of anatomy through multiple-choice questions(MCQ) tests. Also, Laboratory work examination includes dissecting preserved specimens, microscopic study, and computer simulations in objective structured clinical examination (OSCE) situations to reveal their competence in the anatomy courses [28, 36].\",\n \"We used purposive and snowball sampling with maximum variation. Accordingly, we selected 24 medical students from year 1 to senior interns and 12 FMs within a wide range of academic rank and experience from 9 Iranian Universities of Medical Sciences in 2020–2021.\\nThe inclusion criteria for medical students were studying in the first year of basic science to senior interns based on their willingness to participate. The inclusion criteria for faculty members in the anatomy or medical education field were a minimum of five-year teaching experience and a desire to participate. The exclusion criterion was the unwillingness to participate in the study. Initially, we collected data from a medical teacher well known for his high-quality teaching; then, we continued data gathering from medical students, anatomy faculty members, and medical education experts until data saturation was obtained and when no new code was obtained during interviews and repetition of the previous categories and codes [37–40]. Sampling and data coding continued till data saturation when no new code was obtained during interviews and repetition of the previous categories and codes.\",\n \"We utilized semi-structured interviews. At first, we contacted participants by telephone to explain the purpose of the study and research questions. The interviews were done in a quiet place, at an appropriate time, face-to-face, and individually. The interview questions focused on participants’ experiences of learning and teaching anatomy. The interviews started with questions as follows:\\n\\nWhat is your description of successful teaching/learning in an anatomy course?Tell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\n\\nWhat is your description of successful teaching/learning in an anatomy course?\\nTell me about your experiences in learning/teaching anatomy and different situations, including the dissection laboratory.\\nThen, according to the participants’ answers, we asked exploratory questions. We searched for signs of success in teaching and learning, the factors that lead to failure, and the characteristics that can change the behavior and vision of medical students. Each interview took 25 to 80 min, with an average of 45 min.\\nMoreover, we gathered some field notes from the laboratory’s educational atmosphere. Data were collected and analyzed using Microsoft One Note2010, Microsoft, Redmond campus, US.\",\n \"We listened to each recorded interview to get an overall understanding. Then we analyzed verbatim and in-depth using inductive content analysis before the following interview, so each interview guided the next.\\nThe meaning units consisted of words and sentences abstracted and labeled with codes. Twenty-six participants were interviewed in one session, and two or three sessions were conducted for six subjects.\\nThe thematic content analysis includes six stages: [1] familiarizing with data, [2] generating the initial codes, [3] searching for themes, [4] reviewing the themes, (the various codes were compared based on similarities and differences in meaning and were categorized together) [5] defining and naming themes, and [6] preparing the report [41, 42].\\nAfter conducting iterative and comparative line-by-line coding procedures, we sorted similar codes into subcategories. Finally, we rearranged similar subcategories and domains into major categories [39].\\nIn qualitative research, data quality and accuracy are used instead of validity and reliability. Lincoln and Guba have proposed four methods for validating data that many qualitative researchers have used: credibility, dependability, confirmability, and transferability, which will be discussed later [43, 44].\\nWe used various methods to validate the study, such as prolonged involvement (15 months from July 2020 to Nov 2021) with the participants, data, and subjects. Additionally, we performed member checks and expert checks for coding. For the member check, we sent the results of each analysis to the very participant, who would confirm them based on the interview and their own experiences.\\nThe analysis results and categories were shared, approved by the supervisors, and approved by an experienced qualitative researcher.\\nTransferability is defined by the generalizability of the results to fully describe the subject area and the characteristics of the participants with maximum variation in sampling. The generalization of the results was left to the reader to decide on the given information [45].\",\n \"[SUBTITLE] Participants’ characteristics [SUBSECTION] In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\\n\\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\\nstudents\\n\\nGender\\nNoPercent (%)\\nMean Age (y ± SD)\\nMale1562.522.3 ± 8.4Female937.5\\nDegree\\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\\n\\nDemographic Characteristics of the Study Participants\\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\\n\\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\\nBlacksmith Approach\\n\\nRedesigning Educational Mindset (Making a New Forge)\\n\\nCreating a Vision for Students\\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\\nMindful Clinical View\\n\\nRole Modeling and Support (Heating up the Students’ Hearts)\\n\\nJoyful Teaching (Telling Stories/Games/ Humor)\\n√ every trick must be done√ enjoy class√ follow me in the classroom\\nStudent Support System\\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\\n\\nStudy Guide\\n√ Preparedness√ Road map\\nReference Autonomy\\n√ self-learning√ self-direction\\nNeuroanatomical Engagement\\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\\nMindful Dissection, Better Transition\\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\\nDrawing\\n√ easy to learn√ Draw a lot\\nProblem-Based Learning\\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\\nPeer-Assisted Learning\\n√ Peer to the peer group√ Discussing the problem\\n\\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\\n√ Anatomy is hard, bulky, should be taken seriously\\n√ I can help myself more than anyone\\n√ integrated curriculum\\n√ self-study / Self-confidence\\nscarcely happens in the real world!\\n√ professional identity\\n√ prospective physicians\\n√ every trick must be done\\n√ enjoy class\\n√ follow me in the classroom\\n√ susceptible students\\n√ assistance\\n√ Vulnerable\\n√ right supporters\\nSitting and talking to them\\n√ Preparedness\\n√ Road map\\n√ self-learning\\n√ self-direction\\n√ Visual Learning\\n√ Auditory Learning\\n√ Sensory Learning\\n√ Brain Hemispheres\\nLong-term Memorization\\n√ It smells so bad, but we need the observation\\n√ We got used to it\\n√ ethically important\\n√ respect the dead body\\n√ respect to the living body\\n√ easy to learn\\n√ Draw a lot\\n√ Look at the patient’s family\\n√ Deep worry\\n√ What is the next step of the problem?\\n√ Peer to the peer group\\n√ Discussing the problem\\nIn this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\\n\\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\\nstudents\\n\\nGender\\nNoPercent (%)\\nMean Age (y ± SD)\\nMale1562.522.3 ± 8.4Female937.5\\nDegree\\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\\n\\nDemographic Characteristics of the Study Participants\\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\\n\\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\\nBlacksmith Approach\\n\\nRedesigning Educational Mindset (Making a New Forge)\\n\\nCreating a Vision for Students\\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\\nMindful Clinical View\\n\\nRole Modeling and Support (Heating up the Students’ Hearts)\\n\\nJoyful Teaching (Telling Stories/Games/ Humor)\\n√ every trick must be done√ enjoy class√ follow me in the classroom\\nStudent Support System\\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\\n\\nStudy Guide\\n√ Preparedness√ Road map\\nReference Autonomy\\n√ self-learning√ self-direction\\nNeuroanatomical Engagement\\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\\nMindful Dissection, Better Transition\\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\\nDrawing\\n√ easy to learn√ Draw a lot\\nProblem-Based Learning\\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\\nPeer-Assisted Learning\\n√ Peer to the peer group√ Discussing the problem\\n\\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\\n√ Anatomy is hard, bulky, should be taken seriously\\n√ I can help myself more than anyone\\n√ integrated curriculum\\n√ self-study / Self-confidence\\nscarcely happens in the real world!\\n√ professional identity\\n√ prospective physicians\\n√ every trick must be done\\n√ enjoy class\\n√ follow me in the classroom\\n√ susceptible students\\n√ assistance\\n√ Vulnerable\\n√ right supporters\\nSitting and talking to them\\n√ Preparedness\\n√ Road map\\n√ self-learning\\n√ self-direction\\n√ Visual Learning\\n√ Auditory Learning\\n√ Sensory Learning\\n√ Brain Hemispheres\\nLong-term Memorization\\n√ It smells so bad, but we need the observation\\n√ We got used to it\\n√ ethically important\\n√ respect the dead body\\n√ respect to the living body\\n√ easy to learn\\n√ Draw a lot\\n√ Look at the patient’s family\\n√ Deep worry\\n√ What is the next step of the problem?\\n√ Peer to the peer group\\n√ Discussing the problem\\n[SUBTITLE] Central theme: blacksmith approach [SUBSECTION] According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\\n\\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\n\\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\nAccording to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\\n\\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\n\\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\n[SUBTITLE] Sub theme1: redesigning educational mindset (making a new forge) [SUBSECTION] Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\\n\\nCreating a Vision for Students.\\n\\nCreating a Vision for Students.\\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n\\n2)Mindful Clinical View.\\n\\nMindful Clinical View.\\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\nCreating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\\n\\nCreating a Vision for Students.\\n\\nCreating a Vision for Students.\\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n\\n2)Mindful Clinical View.\\n\\nMindful Clinical View.\\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\n[SUBTITLE] Subtheme 2: role modeling and student support (heating up the students’ hearts) [SUBSECTION] The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\nThe professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\n[SUBTITLE] Joyful teaching (telling stories/games/humor) [SUBSECTION] According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\nAccording to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\n[SUBTITLE] Student Support System [SUBSECTION] Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\nBoth faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\n[SUBTITLE] Subtheme 3: sledgehammer approach (thorough engagement in the age of pandemics) [SUBSECTION] While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\\nWhile anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\\n[SUBTITLE] Study guide [SUBSECTION] Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\nMost participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\n[SUBTITLE] Neuroanatomical engagement [SUBSECTION] Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\nParticipants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\n[SUBTITLE] Reference autonomy [SUBSECTION] According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\nAccording to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\n[SUBTITLE] Mindful dissection, better transition [SUBSECTION] The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\nThe first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\n[SUBTITLE] Drawing [SUBSECTION] Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\nTeachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\n[SUBTITLE] Problem-based learning [SUBSECTION] Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\nStudents will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\n[SUBTITLE] Peer-assisted learning [SUBSECTION] Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\\nStudents stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\",\n \"In this study, semi-structured interviews were performed with 24 medical students from year 1 to senior interns, 15 (62.5% were males), 9 (37.5% were females), and 12 faculty members, 6 (50% were males), 6 (50% were females), with various academic ranks and experience from 9 (Shiraz, Tehran, Fasa, Jahrom, Bandarabas, Yasuj, Gerash, Larestan, and Bushehr) Iranian medical schools. The participants’ age ranged from 19 to 62 years old, with a mean of 33.4 ± 8.4 years. Based on the results, 12 students (50%) were in the clinical training period as students, externs, and interns, and 12 participants (50%) were in the basic sciences period. Accordingly, five participants (20%) had a history of being top students, and four (16.7%) had a history of dropout in anatomy courses. Furthermore, the work experience of the faculty members ranged from 5 to 35 years in teaching anatomy; 9(75%) anatomy faculty members and 3(25%) experts in medical education; faculty member participants, 3(25%) were single, and 9 (75%) were married. (Table 1)\\n\\nTable 1Demographic Characteristics of the Study ParticipantsParticipantsCharacteristics\\nstudents\\n\\nGender\\nNoPercent (%)\\nMean Age (y ± SD)\\nMale1562.522.3 ± 8.4Female937.5\\nDegree\\nFaculty membersassistant professors,32537.4 ± 11.7associated professors541.67full professors433.34\\n\\nDemographic Characteristics of the Study Participants\\nFrom all the interviews, 334 codes were extracted. After the dismissal of similar codes and their integration, the experience of medical students was conceptualized into a central theme called the blacksmith approach, which included Three sub-themes: (1) making a new forge (Adequate preparation and mindful beginning), (2) heating the students’ hearts (considering supporting systems that learners need) and (3) using a Sledgehammer’s approach (teaching Anatomy thorough engaging all neuroanatomical regions in the Age of Pandemics). All the concepts were related to each other and resulted in a pattern revealing the experience of students and faculty members on strategies for successful learning and teaching of anatomy lessons. The main themes, as well as subordinate themes, are listed in Table 2.\\n\\nTable 2The Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codesMain ThemesSub Themescodecods\\nBlacksmith Approach\\n\\nRedesigning Educational Mindset (Making a New Forge)\\n\\nCreating a Vision for Students\\n√ Anatomy is hard, bulky, should be taken seriously√ I can help myself more than anyone√ integrated curriculum√ self-study / Self-confidencescarcely happens in the real world!√ professional identity√ prospective physicians\\nMindful Clinical View\\n\\nRole Modeling and Support (Heating up the Students’ Hearts)\\n\\nJoyful Teaching (Telling Stories/Games/ Humor)\\n√ every trick must be done√ enjoy class√ follow me in the classroom\\nStudent Support System\\n√ susceptible students√ assistance√ Vulnerable√ right supportersSitting and talking to them\\nSledgehammer Approach (Thorough Engagement in the Age of Pandemics)\\n\\nStudy Guide\\n√ Preparedness√ Road map\\nReference Autonomy\\n√ self-learning√ self-direction\\nNeuroanatomical Engagement\\n√ Visual Learning√ Auditory Learning√ Sensory Learning√ Brain HemispheresLong-term Memorization\\nMindful Dissection, Better Transition\\n√ It smells so bad, but we need the observation√ We got used to it√ ethically important√ respect the dead body√ respect to the living body\\nDrawing\\n√ easy to learn√ Draw a lot\\nProblem-Based Learning\\n√ Look at the patient’s family√ Deep worry√ What is the next step of the problem?\\nPeer-Assisted Learning\\n√ Peer to the peer group√ Discussing the problem\\n\\nThe Blacksmith Approach: Best strategy for teaching and learning in the medical anatomy course. Themes, sub-themes, and codes\\n√ Anatomy is hard, bulky, should be taken seriously\\n√ I can help myself more than anyone\\n√ integrated curriculum\\n√ self-study / Self-confidence\\nscarcely happens in the real world!\\n√ professional identity\\n√ prospective physicians\\n√ every trick must be done\\n√ enjoy class\\n√ follow me in the classroom\\n√ susceptible students\\n√ assistance\\n√ Vulnerable\\n√ right supporters\\nSitting and talking to them\\n√ Preparedness\\n√ Road map\\n√ self-learning\\n√ self-direction\\n√ Visual Learning\\n√ Auditory Learning\\n√ Sensory Learning\\n√ Brain Hemispheres\\nLong-term Memorization\\n√ It smells so bad, but we need the observation\\n√ We got used to it\\n√ ethically important\\n√ respect the dead body\\n√ respect to the living body\\n√ easy to learn\\n√ Draw a lot\\n√ Look at the patient’s family\\n√ Deep worry\\n√ What is the next step of the problem?\\n√ Peer to the peer group\\n√ Discussing the problem\",\n \"According to the participants, in the student’s learning and maturity process, the teacher and the student act rationally and consciously and are aware of the factors that motivate and control them. Thus, they accept the risks, specific conditions, and hardships that must be tolerated. The student should handle this hardship to acquire a capacity to change, and the teacher should facilitate the change and their primary goal of teaching. In a sense, this is the same approach as the Blacksmiths. Blacksmithing symbolizes endurance, honor, and work under challenging conditions in our culture. (Fig. 1) We took the central theme from an interview with a professor who said:\\n\\nFig. 1The role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\\n\\nThe role of the blacksmith approach in the stages of teaching and learning anatomy in medical students is based on the study’s results. First, the blacksmith (educational system) must prepare a new blacksmith (redesign the educational mentality and create a new perspective on anatomical knowledge in students) and prepare a new forge; then, he/she must heat the metal (supporting young students and having good role-modeling) and then shape the metal with a sledgehammer (the sledgehammer approach to using different teaching strategies and methods) to create a sword or self-confident and knowledgeable learner in medicine (professionalism)\\n”…The Sledgehammer should be heavy. Its nature is heavy; otherwise, it will not be capable of smoothing any iron. The nature of some courses, such as anatomy, is challenging, but we must make it enjoyable for the student….“ (FM no.6).\",\n \"Creating an educational context is the first step for young students to begin their transition in the learning process. This context was found to have two aspects as follows:\\n\\nCreating a Vision for Students.\\n\\nCreating a Vision for Students.\\nSome students mentioned a change in their mindset from the beginning with the idea that they have a difficult task ahead and should know what is helpful or not. In addition, realizing that the student is the most important person to help themselves seemed to provide a better sense of success.\\nAccording to the participants, anatomy is one of the most voluminous and complex courses in medicine, which students should take seriously from the beginning.“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\n“Anatomy is hard, bulky, and should be taken seriously.“ (Student no. 9)\\nEven though many have oversimplified it, successful students have attempted to learn it through vision and self-awareness.“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\n“… After entering the school, my mentor advised me to study anatomy as soon as possible. An anatomy lesson is not something you would say I will study and comprehend the concepts overnight or for a week. You should attend the classes regularly. Students should provide accessible learning resources and not seek shortcuts…” \\\" (Student no. 1).\\nParticipants believed that successful students have good self-confidence and leadership in their learning process. Some students emphasized that the teachers can first build self-confidence and make them self-centered and self-regulated in learning. This concept requires the teachers to listen to them.“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\n“... If I realized the concepts satisfactorily, I would get a lot out of my effort. First, I accepted that anatomy is complicated, but I can help myself more than anyone. I started self-study. More action boosts self-confidence, which is required to become a doctor.“(Student no. 6)\\nFinally, the participants emphasized the importance of integration. Integration is simply defined as merging relevant content or subject areas. Teachers mention that the goal is to integrate and create a general perspective rather than a fragmented perspective of concepts. Some students believed that while multidisciplinary integration was provided, professors’ art of education, such as mastery of the teaching and learning process, generated the necessary harmony between the content presented. If well-coordinated, all aspects mutually reinforce each other, promoting learning in an integrated way.“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\n“... I enjoyed our integrated curriculum because that led to a better understanding of medicine. For example, discussing the anatomy of the gastrointestinal system gave me invaluable and integrated knowledge, and I enjoyed it ... “(Student no. 4)\\nAnatomy professors acknowledged that integration with clinical courses had changed their perspective, reflected in their teaching.“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n“…During the integration process and while I was listening to the viewpoints of the clinical colleagues, I realized what I thought to be the crucial thing that students should all be aware of scarcely happens in the real world!...“ (FM no. 2)\\n\\n2)Mindful Clinical View.\\n\\nMindful Clinical View.\\nMindfulness is defined as awareness of what is happening at the present moment. Some professors believed that they must be aware of the ultimate goal while teaching at each moment. Therefore, mindfulness is influenced by one’s dreams.“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\n“…I try to teach anatomy in a way that our current students and prospective physicians can diagnose heart enlargement in the future…” (FM no. 7)\\nIf anatomy teachers gain a conservative view, the students can achieve integrity and professional identity as doctors.“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\\n“... The student’s learning and professional identity depend largely on knowing and comprehending anatomy. Teachers can play the role of a good teacher in any situation, whether a crisis (like the Covid-19 pandemic) or not. A good teacher always applies the right approach....“ (FM no. 1).\",\n \"The professors’ experiences in educating successful students during anatomy courses show that they should reduce unnecessary complaints and criticisms to penetrate their hearts and light the torch of knowledge in their minds. Feelings for the learners should be expressed, and love for them should be shown. There should be a role model for them in academia. If the love of science enlightens students, they are strengthened to endure hardships and problems. One of the professors said in this regard:“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\\n“…As an ancient Persian anecdote, “to attract the heart of a friend, we must do every trick.“ I sat down during soccer World Cup games and watched part of the game with the students…” (FM no. 8)\",\n \"According to the participants, turning hard work into enjoyable teaching and learning is the art of medical teachers. A few faculty members believe that humor and mixing lessons with fun and games can make students aware of their superior position as a professor. They find icebreaking essential at the beginning of each discussion.“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\n“…If students enjoy the class and assume that learning is a plausible activity, they will not waste their time. Most of the wasted time and students’ naughtiness occur because of the attractiveness of the classroom and teacher….“ (FM no. 10)\\nSuch an approach strengthens the bond between the students and teachers and leads to effective role modeling.“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\n“... If the student does not follow me in the classroom, what is the benefit of my teaching class?“ (FM no..8)\\nStudents can understand whether their teacher can be a trusted and knowledgeable role model and respond better if they encounter a fun and supportive teacher.“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\\n“…good teachers are fully aware of all their behavior and movements and know what to teach in the classroom. They avoid traditional teaching methods and teach enjoyable for both students and teachers. Students will not be tired and enjoy the learning process….“ (Student no. 10)\",\n \"Both faculty members and students mentioned the student support system as influential. All the participants believe that student support is helpful for vulnerable students, those with learning disabilities, and successful and elite students. The faculty members referred them to counseling, helped them solve their educational and emotional problems, and supported their families emotionally and financially.“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\n“…Student support is essential, but it is not just for high-risk and susceptible students. Sometimes elites need even more assistance, and I have seen and experienced how sometimes talented kids are vulnerable...“ (FM no.8).\\nSuch support, mainly when anatomy professors are referred to as the first role models for medical students, motivates students to improve over the past.“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\\n“…I had the support of the family and could work harder and enjoy learning; they did not hesitate to do anything for me. The teachers and faculty members were sometimes right supporters and solved my problem….“ (Student no. 3)\",\n \"While anatomy is perceived as one of the most challenging courses for preclinical students, the advent of the Covid-19 pandemic adds to the challenge. Successful anatomy teachers state that they must apply a combination of educational theories quickly, seamlessly, and effectively in the classroom or online. They believe that teaching anatomy in pandemic situations should be accompanied by joy and vivacity and increase the moments of discovery and intuition for learning. The faculty members seem to struggle to get the most involvement from students.\\nTeaching approaches that trigger student involvement include a study guide, reference autonomy, neuroanatomical engagement, drawing, problem-based learning, and peer-based learning.\",\n \"Most participants believe preparedness is an essential strategy for success in learning and teaching. The best way is to provide a friendly study guide. The study guide is like a travel guide designed by an instructor to guide learners. It can increase the student’s insight into the subject. Some professors had the experience of contributing to making “road map” books for anatomy. They believe that such a study guide is also a studying resource. Others found the study guide essential for managing the students’ learning.“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\n“... Student guidance is an important issue that needs to be planned and implemented. We need more insight and hard work in pandemic conditions. Our “road map” book that primarily serves as a reference of anatomy is a study guide as well…” (FM no. 3)\\nStudents assume that having a study guide is an insightful experience with more potential than it seems. They argue that a good stud guide must include common mistakes and guide them in all learning experiences, especially in cadaver dissection sessions.“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\\n“... The study guide helps a lot in learning, and it is excellent if students’ common mistakes are made in this lesson, for example, in the lab and when we dissect the cadaver ...“ (FM no. 6)\",\n \"Participants believed that people learned new abstract concepts better when presented in visual and auditory ways. Professors believed that due to the functions of the brain hemisphere, the information process occurred through two separate systems in the brain. The right half is dedicated to the image, and the left half is word processing. Visual learning complements auditory learning. Sensory learning, which happens by working on the cadaver, complements student learning. Such a multidimensional approach leads to long-term memorization.“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\\n“…I always use fascinating slides and video clips. If necessary, I explain my experience through the film, which enhances the understanding and facilitates long-term memorization…” (FM no. 5)\",\n \"According to the participants, appropriate resources play an essential role in learning. A good resource should encourage the students to become more active and reinforce their thinking, reasoning, and comprehension. Resources must be dynamic to adapt in times of crisis (pandemic) and change to meet the needs of 21st-century learners using current technologies. There seems to be a demand for respecting autonomy in both student and teacher experiences. They react better when students feel free and supported to choose their resources.“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\n“…For anatomy, I use Gray’s Anatomy book. However, recently, one of my professors introduced Moore’s Anatomy book. I felt this book was better in the field of …” (Student no. 2)\\nTeachers argue that if provided with a dynamic curriculum design that encompasses clear goals, they can select from all the available resources and provide better ones. They believe that students should be taught to search and find valuable resources when they are highly interested in topics.\\nWhen teachers accept the principle of self-learning and self-guidance, they listen to their students feel more confident, leading to more enthusiasm and thus creating future lifelong learners.“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\\n“…I had a student who came up to me and asked me to explain how to learn anatomy. I was surprised because I was unfamiliar with the question. I told him to get two books, look at them, and return them to me the day after. He came while he got many valuable points, and now he is looking for further resources. Now, he is one of my most successful students, and one of my most important duties is introducing excellent resources to students...“ (FM no. 4)\",\n \"The first official and scientific encounter of young students with the human body occurs in dissection sessions. The first sessions familiarize the students with the disgusting aspects of gross medical procedures in which teachers play a facilitating role in such transition.“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\n“…I did not feel good on the first day, and many of my friends could not see the cadavers. It smelled so bad, but when we got used to it and started learning, we had better feelings, especially when we had a practical examination …” (Student no.14)\\nAnatomy teachers are the game changers who can seed respect towards the human body in students’ minds.“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\\n“… A cadaver is the most important teaching tool in the anatomy course and serves as a teacher’s blackboard. It is full of teaching points for medical students. I think this is ethically important as well. They should learn to respect the dead body highly to show respect to the living body in the future….“ (FM no.12)\",\n \"Teachers and students have valuable experiences in learning and memorizing through drawing. Drawing helps them to understand the topic, consolidate what they have learned, and remember it in long-term memory.“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\n“…Anatomy has a regional nature. It means that, for instance, everything in the chest area should be learned simultaneously and as a whole. I use drawing to show the adjacency of structures so that students can better learn and consolidate the topic...“ (FM no. 3)\\nFor some learners, drawing becomes a ritual; one of the talented students said:“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\\n“…I always tell other students to draw pictures to understand anatomy correctly. Draw a lot! Anatomy cannot be understood without painting. No matter how strong you are in abstract thinking, still, draw! I had a notebook in which I was painting...“ (Student no.5)\",\n \"Students will feel more satisfied if they are involved in future career problems. By using this approach, students are also emotionally involved in future challenges and are taught to enjoy solving future issues. Those problems that include detailed scene descriptions in a socio-emotional context and are related to cultural, moral, and social values seem to engage students better, help long-term memorization, and facilitate transition into future doctors.\\nOne of the professors said:“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\\n“… I will show the students a picture of the patient and tell them that the person has subarachnoid hemorrhage. Look at the patient’s family. You can see their deep worry. Tell me where the bleeding source is and what is the first step?... \\\" (FM no. 6)\",\n \"Students stated that learning in small groups was attractive because they acquired interpersonal skills like listening, speaking, discussing, and group leadership. Besides, it promotes higher cognitive abilities such as reasoning and problem-solving.“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\\n“...For me, the best learning tool is a peer to peer group work. I have good experience in all subjects from literature to anatomy utilizing this method…” (Student no.5)\",\n \"We first scrutinized the experience of teaching and learning anatomy through qualitative analysis. Next, we conceptualized a new approach to help medical students in anatomy courses have more satisfactory learning outcomes and a better transition into resilient future doctors. We called it “the Blacksmith Approach.“ building a sword (prosperous and resilient learner) takes three steps. 1. Creating a new forge (adequate preparation and mindful beginning) 2; heating students’ hearts (support and role modeling) 3. Sledgehammer’s approach (thorough engagement in the age of pandemics). (Fig. 1)\\nWhile other proposed educational methods in teaching anatomy focus mainly on learning outcomes during anatomy courses [14–19], the blacksmith approach tries to reach two goals, i.e., better teaching-learning outcomes and helping young students transition into future medical professionals. In fact, this is not another teaching method. It best serves as an educational framework for any pivotal, preclinical course capable of helping students acquire new roles and tackle challenges. In this sense, medical educators can adopt all prior teaching techniques as long as they help students reach better learning outcomes or have better transition experiences.\\nThe blacksmith approach can be an educational implication of Schlossberg’s 4 S model for transition. Transition is a process that enables individuals to incorporate change into their life. The 4 S theory argues that an individual goes through a cyclic experience of 4 stages during the transition, i.e., situation, self, support, and strategies [46]. We encounter two types of transition in an educational setting. One happens within the learners and one within the educators. Although more attention is put on the former, we cannot achieve desired outcomes if we neglect the latter. While in the 4 S model, these four stages happen within the individual, the Blacksmith Approach creates a compatible, three-step learning environment for students and teachers to transition by creating a new attitude and mindset for both groups. As Browne et al. stated, individual motivation, background and circumstances, a sense of control, organizational support, effective networking, and information-seeking behavior contributed to a successful transition into and maintaining a strong self-identity as a medical educator [47].\\n[SUBTITLE] First Step: A new forge (mindset) [SUBSECTION] must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\\nmust be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\\n[SUBTITLE] Second step [SUBSECTION] The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\\nThe students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\\n[SUBTITLE] Third step: Sledgehammer’s approach [SUBSECTION] We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.\\nWe found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.\",\n \"must be built for the students. This step is in line with the first stage of the four S model, i.e., Situation awareness. Here, the students become aware of what is happening. The teachers here have two roles. First, they should deliver the following messages to the students: (1) you are supposed to actualize a new identity (role change), (2) anatomy (as a pillar of medicine) provides a problematic task ahead (stress), and (3) you can help yourself more than anyone else. (control) Furthermore, the students should try self-study to increase their self-confidence (trigger). Secondly, the teachers should acquire a mindful clinical standpoint, i.e., developing good clinical knowledge and adopting active learning methods based on teamwork and clinical scenarios that encourage active participation and self-directed learning as if they are future doctors. This strategy has been shown to reduce student stress [11, 13] and help them react better as future physicians [3, 8, 48].\",\n \"The students’ hearts must be warmed up through support and role modeling. This step is compatible with the following two stages in the 4 S model, i.e., self and support. The students evaluate their demographic and psychological characteristics, and then they need to be aware of support systems and be able to get help [27, 29]. Here again, the role of anatomy teachers is critical. As medical students see anatomy as a gate to medicine, the anatomy teacher can be known as the gatekeeper. They will inevitably actualize their teachers as wise role models who are righteous supporters. The anatomy faculty member should be aware of their role modeling. Role modeling can help students foster resilience for future change, which leads to professionalism and integrity [7, 18, 49–52]. The teachers should adopt ice-breaking methods, use their sense of humor, tell stories, design games, and have friendly chats with students outside the classroom. In addition, anatomy teachers should gain the skills and knowledge for counseling and get familiar with support strategies to become prominent supporters of students [53–56].\",\n \"We found that, currently, preclinical students are facing two overwhelming stressors. On the one hand, the covid crisis has detached the students from their teachers and resources [15, 18, 57]. On the other hand, students need to handle their transition into knowledgeable medical professionals. Therefore we need to incorporate teaching methods that assist our detached students in coping with their stressors. The last stage in the 4 S model also deals with coping strategies. Students may use several coping strategies like information seeking, selective ignoring, and reframing the challenge [58].\\nAccordingly, the third step in our model includes three components; thorough engagement, reframing the situation, and selective information seeking. Regarding the first component, we encourage meticulously choosing more active and engaging learning techniques. Problem-based learning and Peer-based discussions in small groups, drawing, mind mapping, body painting, using play-doh, and interactive video content, are advised. According to Pandey and Zimitat (2007), a practical approach to anatomy education positively correlates with teaching and reflective learning quality. Successful anatomy learning requires balancing memorization with understanding and visualization[59]. Insightful experiences in teaching anatomy with this strategy are found in Australia (head and neck course) [15] and India (musculoskeletal anatomy) [13]. As Goodman et al. (2006) noted in their previous book, those with higher self-esteem and mental mastery are better able to cope with challenges. [60] A successful learning experience gives students a sense of control over their course, while applying a study guide, reference autonomy, and support system can help them gain self-esteem [61–63].\\nReframing the situation, As the second component, is where the students need to redefine their roles and activities. Two sets of strategies can help students to reframe. 1) Using problem-based and peer-assisted learning with a flavor of sociocultural contexts can prepare students for better learning in a positive, non-intimidating learning environment to redefine themselves as future doctors for diagnosing and curing patients. Socio-emotional context and bringing drama elements into case scenarios have been recently emphasized to induce resilience and professionalism in medical students (2.14, 64).\\nRecent evidence indicates that gamification leads to positive learning outcomes, higher motivation, and engagement. However, several downsides should also be considered, including increased competition and task evaluation difficulties. Besides, it depends highly on the context in which the gamification is being implemented and its users [65–68]. Hamari et al. (2014) have reviewed empirical literature, showing that gamification has proven effective in many fields, most notably education. Studies in education and learning contexts illustrate how the gamification of learning outcomes has been positive by increasing motivation and engagement in learning and enjoyment tasks over time. However, the studies also highlighted some negative points that require attention, such as the effects of increased competition, task evaluation difficulties, and design features. The review indicates that gamification provides positive effects; however, the effects greatly depend on the context in which the gamification is being implemented and the users who use it [66]. Recent studies suggest that joy and humor facilitate a person’s transition into more socially severe roles (like becoming a doctor) [34]. Hence, teachers should use joyful techniques such as painting, ice braking, storytelling, gamification, dance, and music [4, 9, 15, 53].\\nThe last component deals with selective information seeking. Students should have a study guide and be free to choose their desirable resources (anatomy reference books, digital resources, or web-based social media videos). We advise the medical school to design a dynamic curriculum encompassing clear goals. In this setting, the study guide can act as a tool for transferring core knowledge in anatomical sciences. As a result, the tedious aspect of bulky content in anatomy courses can be avoided, and the educational efficiency of synchronous and asynchronous learning can be increased in the pandemic era [55, 56, 61, 69, 70].\\nThis research is in line with constructive learning theory that emphasizes student-centered learning, which states that students should be actively involved in education and contribute to knowledge creation [71]. In a special effort to create a modern educational curriculum at Stanford University, Prober et al. recommended building a framework of core knowledge, embedding it into rich interactive formats, and encouraging students to pursue knowledge in some deep, but not all, domains [63]. The Blacksmith approach facilitates personal and in-depth pursuits that help transfer students into lifelong learners and adapt to challenging situations [63, 64]. This statement is consistent with Wald and Monteverde (2021). They report that educational resilience is vital in supporting and sustaining healthcare professional identity development and facilitating the progress of students’ moral resilience and complexity, especially in times of pandemic [25].\\nAccording to the participants’ viewpoints, the applicability of the blacksmith approach is one of the most important aspects of improving the quality of anatomy teaching and learning in any situation, such as current critical conditions.\\nOur university’s online programs were adequately integrated into the curriculum design and used to guide students on a learning journey rather than merely as another resource; perhaps, these tools could be beneficial. However, in such circumstances, the epidemic and sanctions of Iran and the lack of resources, students and professors should take measures to optimize education to facilitate the teaching and learning of this course. Accordingly, from the above point, through a qualitative study, it was assumed that there were signs of successful teaching and learning and internal change in deep search among students and faculty members.\\nRevolution in medical education is a global program, and many medical schools have perceived investigation and experienced challenges [63, 64]. Correspondingly, an active and integrated strategy for teaching and learning anatomy has been emphasized in other medical education courses and areas, including a contemporary medical curriculum. Managers, practitioners, and professors can use the elements of this strategy to increase the efficiency of medical education, improve students’ knowledge, skills, and attitudes, and ultimately maintain and promote understanding of anatomy and other courses. Also, from the participant’s perspective, the applicability of students learning experiences and faculty members’ teaching experiences in the context and workplace conditions was one of the factors that changed the view of anatomy as an introductory course in medicine.\",\n \"As this study was conducted in one country, it is suggested that more studies and insights be obtained from other participants to increase the generalizability of findings. Besides, further work should be done to provide more practical guidance in the educational system regarding the transition of medical students in teaching and learning anatomy in the current pandemic crisis.\",\n \"Medical students experience a fundamental evolution into responsible and lifelong learners qualified as professional doctors. Educational systems focus primarily on teaching and learning, while students’ transition can be facilitated by a three-step model called the Blacksmith Approach, based on the study results. Initially, the blacksmith (educational system) must prepare a new forge. It means anatomy teachers should redesign the academic mentality of students by delivering messages about role change, difficulties ahead, and the importance of self-guidance, along with implementing clinical scenarios in active learning methods. Then, he/she must heat the metal. It implies that anatomy teachers, as gatekeepers of medicine, should adopt amusing teaching methods while embracing their role modeling and actively applying support strategies. Finally, he/she may shape the metal with a sledgehammer. Teachers need to use three components to create a sword (self-confident, accountable, resilient professional); thorough engagement by using active and engaging teaching techniques, reframing the situation with problem-based clinical scenarios, and joyful teaching. And selective information seeking via study guide and reference autonomy.\\nIn conclusion, an active and integrated strategy for teaching and learning anatomy is suggested to be applied in other courses and areas of medical education in the current medical curriculum.\",\n \"[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\nBelow is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction",null,"materials|methods",null,null,null,null,null,"results",null,null,null,null,null,null,null,null,null,null,null,null,null,null,"discussion",null,null,null,null,"conclusion","supplementary-material",null],"string":"[\n \"introduction\",\n null,\n \"materials|methods\",\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n null,\n \"conclusion\",\n \"supplementary-material\",\n null\n]"},"keywords":{"kind":"list like","value":["Anatomy education","Teaching","Learning","Medical faculty","Medical students","Qualitative"],"string":"[\n \"Anatomy education\",\n \"Teaching\",\n \"Learning\",\n \"Medical faculty\",\n \"Medical students\",\n \"Qualitative\"\n]"}}},{"rowIdx":375,"cells":{"title":{"kind":"string","value":"Prevalence and factors associated with postpartum pelvic girdle pain among women in Poland: a prospective, observational study."},"pmid":{"kind":"string","value":"36266709"},"background_abstract":{"kind":"string","value":"Pelvic girdle pain (PGP) is a type of pregnancy-related lumbopelvic pain. This study aimed to examine the prevalence, severity, and factors associated with postpartum PGP in a selected group of postpartum women in Poland."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"This was a prospective, observational study. In phase 1, 411 women were recruited 24-72 h postpartum. The prevalence of PGP was assessed by a physiotherapist using a series of dedicated tests. Pelvic floor muscle function and presence of diastasis recti were assessed via palpation examination. Age, education, parity, mode of delivery, infant body mass, body mass gain during pregnancy, the use of anesthesia during delivery and were recorded. In a phase 2, 6 weeks postpartum, the prevalence of PGP and its severity were assessed via a self-report."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"In phase 1 (shortly postpartum), PGP was diagnosed in 9% (n = 37) of women. In phase 2 (6 weeks postpartum), PGP was reported by 15.70% of women (n = 42). The univariable analyses showed a higher likelihood of PGP shortly postpartum in women who declared PGP during pregnancy (OR 14.67, 95% CI 4.43-48.61) and among women with abdominal midline doming (OR 2.05, 95% CI 1.04-4.06). The multivariable regression analysis showed significant associations in women with increased age (OR 1.12, 95% CI 1.01-1.21) and declaring PGP during pregnancy (OR 14.83, 95% CI 4.34-48.72)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Although the prevalence of postpartum PGP among women in Poland is lower than reported in other countries, it is experienced by almost every tenth women shortly postpartum and every sixth can report similar symptoms 6 weeks later. Age, PGP during pregnancy and abdominal midline doming were associated with experiencing PGP shortly postpartum."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Pregnancy","Female","Pelvic Girdle Pain","Prevalence","Prospective Studies","Poland","Postpartum Period","Pregnancy Complications"],"string":"[\n \"Humans\",\n \"Pregnancy\",\n \"Female\",\n \"Pelvic Girdle Pain\",\n \"Prevalence\",\n \"Prospective Studies\",\n \"Poland\",\n \"Postpartum Period\",\n \"Pregnancy Complications\"\n]"},"pmcid":{"kind":"string","value":"9585777"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"A total of 415 women were invited to participate, 4 of whom did not consent to join the study. Thus, 411 women were included in phase 1 (shortly postpartum). Table 1. presents the characteristics of the participants. In phase 2 (“follow-up”), 268 women replied to the text message (65.2% of the initial group). No statistically significant differences were found between those who responded to the message and those who did not in terms of variables assessed shortly postpartum (in phase 1): BMI, age, presence of PGP postpartum, parity, DRA severity, and ability to activate pelvic floor muscles.\n\nTable 1Characteristics of the study group, n = 411All participantsn = 411With PGP n = 37Without PGP n = 374\nAge\n31.17 ± 4.0132.35 ± 4.3331.06 ± 3.96Height [cm],167.46 ± 5.69167.00 ± 6.64167.51 ± 5.60\nBMI before pregnancy\n22.14 ± 3.5722.46 ± 3.2022.10 ± 3.60\nBody mass gain in pregnancy [kg]\n14.30 ± 4.9414.92 ± 4.2814.24 ± 4.99Education, n (%) †vocational education2 (0.5)11secondary education43 (10.5)242university education366 (89)35331\nParity, n (%)\n1.72 ± 0.961.95 ± 1.101.70 ± 0.941209 (50.80)2143 (34.80)337 (9.00)412 (3.00)58 (2.00)61 (0.20)71 (0.20)\nMode of the last delivery, n (%)\nvaginal387 (94.20)37 (100)350 (93.60)cesarean ‡18 (4.30)018 (4.80)vacuum extractor ‡6 (1.50)06 (1.60)\nPerineal injury during recent delivery, n (%)\nnone162 (39.50)18 (48.60)144 (38.50)1st grade138 (33.50)9 (24.30)129 (34.50)episiotomy111 (27)10 (27)101 (27)\nAnesthesia during last delivery, n (%)\nnone238 (57.91)24 (64.86)214 (57.22)epidural154 (37.47)12 (32.43)142 (37.97)spinal19 (4.62)1 (2.70)18 (4.81)\nInfant body mass ≥ 4000 g\nyes336 (81.75)6 (16.22)330 (88.24)no75 (18.25)31 (83.78)44 (11.76)† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\n\nCharacteristics of the study group, n = 411\n† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\nAmong the patients in phase 1 of the study (shortly postpartum), 47.9% (n = 197) reported PGP symptoms during pregnancy. In phase 1, PGP was diagnosed in 9.0% (n = 37) of women at the early postpartum stage. In phase 2 (6 weeks postpartum), PGP was reported by 15.7% of women (n = 42). Table 2 presents the prevalence and type of PGP across different time points, and Table 3 shows the severity of pain and functional disturbances in both phases of the study.\n\nTable 2Prevalence and types of PGP across the time points.TimepointDuring pregnancyEarly postpartum6 weeks postpartum\nAssessment form\n\nself-reported, retrospective\n\nclinical assessment\n\nself-reported\n\nN\n\n%\n\nN\n\n%\n\nN\n\n%\n\nPGP\n197/41147.9337/411942/26815.70\nPGP\n\ntype\n\nPosterior Pelvic Pain\n86/19743.709/3724.3016/4238.10\nUnilateral pain\n17/1978.602/375.402/424.80\nSymphyseal pain\n44/19722.309/3724.3012/4228.60\nPelvic Girdle Syndrome\n47/19723.9016/3743.2011/4226.20\nUnilateral pain + symphyseal pain\n3/1971.501/372.701/422.40\n\nPrevalence and types of PGP across the time points.\n\nPGP\n\n\ntype\n\n\nTable 3The severity of pain and functional limitations at phase 1 and 2 of the studyTimepointEarlyPostpartum6 weekspostpartum\nMean (SD)\n\n(min-max)\n\nN\n\nMean (SD)\n\n(min-max)\n\nN\n\nNRS\n5.34 (1.74)(3–9)37/374.63 (2.04)(1–8)24/42\nPGQ [%]\n48.87 (16.40)(23.61–82.61)32/3725.80 (14.90)(6.67–56.94)13/42The values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\n\nThe severity of pain and functional limitations at phase 1 and 2 of the study\n5.34 (1.74)\n(3–9)\n4.63 (2.04)\n(1–8)\n48.87 (16.40)\n(23.61–82.61)\n25.80 (14.90)\n(6.67–56.94)\nThe values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\nThe univariable analyses showed a higher likelihood of PGP shortly postpartum in women who declared PGP during pregnancy and among women with doming at the abdominal midline in the projection of linea alba (Table 4.).\n\nTable 4Univariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)PGP (+)n = 37PGP (-)N = 374OR (95% CI)p-value\nPGP during pregnancy (yes)\n\nn (%)\n34 (91.90)163 (43.60)14.67 (4.43–48.61)\n< 0.01\n\nAge\n\nmean (SD)\n32.35(4.33)31.06 (3.96)1.08 (0.99–1.18)0.06\nBody mass gain during pregnancy\n\nmean (SD)\n32.68 (4.60)31.59 (4.37)1.03 (0.96–1.10)0.42\nBMI before pregnancy\n\nmean (SD)\n22.46 (3.20)22.1 (3.60)1.03 (0.94–1.12)0.56\nNumber of previous deliveries mean (SD)\n1.95 (1.10)1.7 (0.94)1.27 (0.93–1.71)0.14\nUrinary incontinence during pregnancy or before\n\n(yes) n (%)\n15 (40.50)205 (54.80)0.56 (0.28–1.12)0.10\nAbdominal midline doming (yes), n (%)\n20 (54.10)134 (36.40)2.05 (1.04–4.06)\n0.04\n\nDR severity\n\nn (%)\n\nnone, IRD < 2\n9 (24.30)149 (40.50)1.27 (0.95–1.69)0.11mild,\nIRD 2;< 3\n8 (21.60)72 (19.60)\nmoderate, IRD 3;<4\n12 (32.40)75 (20.40)\nsevere, IRD > 4\n8 (21.60)72 (19.60)\nOxford scale\n\nMean (SD)\n2.27(0.65)2.28 (0.90)0.99 (0.67–1.45)0.96\nReissing scale\n\nmean (SD)\n0.32 (1.11)-0.53 (0.96)1.25 (0.88–1.77)0.21\nCorrect activation of pelvic floor (yes), n (%)\n13 (35.10)134 (35.80)1.36 (0.69–2.70)0.37\n\nUnivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\n\nPGP during pregnancy (yes)\n\n\nn (%)\n\n\nAge\n\n\nmean (SD)\n\n\nBody mass gain during pregnancy\n\n\nmean (SD)\n\n\nBMI before pregnancy\n\n\nmean (SD)\n\n\nUrinary incontinence during pregnancy or before\n\n\n(yes) n (%)\n\n\nDR severity\n\n\nn (%)\n\nmild,\n\nIRD 2;< 3\n\n\nOxford scale\n\n\nMean (SD)\n\n\nReissing scale\n\n\nmean (SD)\n\nBased on the obtained prevalence of postpartum PGP, we could include up to 4 variables into the multivariable model [38]. The final model, in which we obtained statistically significant results for all included items, consisted of 3 variables. This multivariable regression analysis showed that the odds of having PGP shortly postpartum were higher in women with increased age (10% higher likelihood with every year of age), declaring PGP during pregnancy, and with higher values of Reissing scale. To enhance the interpretability, we then analyzed the Reissing scale in the following categories: hypotonus (range − 3 to -1) and increased muscle tone (range 1 to 3), with the reference value being normotonus (0). However, after this procedure, the Reissing scale became not statistically significant (Table 5). The calculated Nagelkerke R-square was 0.2.\n\nTable 5Multivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)OR (95% CI)p-value\nPresence of PGP during pregnancy (yes)\n14.83 (4.340-48.721)\n< 0.0001\n\nAge\n1.12 (1.009–1.214)\n0.032\n\nReissing scale\n\nReissing scale – not grouped\n1.43 (1.003–2.046)\n0.048\n\nReissing scale when grouped\n\nnormal tone (0)\nreferencereference\ndecreased tone \n\n(range − 3 to -1)\n0.53 (0.227–1.220)0.134\nincreased tone \n\n(range 1 to 3)\n0.37 (0.594–4.078)0.368\n\nMultivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\n\ndecreased tone \n\n\n(range − 3 to -1)\n\n\nincreased tone \n\n\n(range 1 to 3)\n"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"The findings presented in this study suggest that every tenth Polish woman may experience PGP during the first days postpartum and every sixth can report similar symptoms 6 weeks later. The pain intensity and functional limitations tend to subside over time: from moderate pain intensity and functional limitation shortly postpartum to mild/low 6 weeks later. Nevertheless, postpartum PGP should not be ignored, especially in the context of the observed continued increase in chronic pain syndromes and their associated consequences. Older age, PGP during pregnancy, and doming of the abdominal midline at the level of linea alba were associated with the experience of PGP within the first days postpartum. Our study showed no association between pelvic floor function and PGP shortly postpartum. However, this may be due to the chosen methodology (assessment shortly postpartum)."},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Setting","Participants","Procedure","Measures/Variables","PGP prevalence and severity","Pelvic floor measurements","DRA measurements","Statistical analysis","Strengths and limitations","Implications"],"string":"[\n \"Background\",\n \"Methods\",\n \"Setting\",\n \"Participants\",\n \"Procedure\",\n \"Measures/Variables\",\n \"PGP prevalence and severity\",\n \"Pelvic floor measurements\",\n \"DRA measurements\",\n \"Statistical analysis\",\n \"Strengths and limitations\",\n \"Implications\"\n]"},"other_sections_texts":{"kind":"list like","value":["Symptoms of pelvic girdle pain (PGP) are commonly reported to healthcare providers by pregnant or postpartum women. This musculoskeletal disorder is a form of low back pain experienced between the posterior iliac crests and the lower edge of the gluteal folds, most commonly in the vicinity of the sacroiliac joints. PGP also includes pain in the pubic symphysis, occurring in isolation or in conjunction with other pelvic joints. Patients with PGP have reduced tolerance to standing, walking, sitting, and changing positions [1]. Pregnancy-related PGP may appear as early as the first trimester of pregnancy or can be delayed up to 3 weeks postpartum [2]. PGP is associated with significantly more pain and functional limitations than lower back pain [3]. The pain often subsides after delivery, but some women continue to have persistent symptoms postpartum. Among women reporting PGP during pregnancy, 1 in 10 will suffer from it up to 11 years later [4]. This significantly impacts their quality of life [5]. In Poland, the prevalence of PGP during pregnancy was reported by 42% of women [6]. To our knowledge, there are no studies reporting on the prevalence of postpartum PGP in Poland or any other country in the central-eastern region of Europe.\nAssessment of muscle impairments has been recommended in Clinical Practice Guidelines for PGP in the postpartum population [7]. To date, several studies investigated musculoskeletal factors that could be associated with pregnancy-related PGP, including the function of pelvic floor muscles and diastasis rectus abdominis (DRA). Coordination between lumbopelvic and abdominal muscles and fascia was suggested to play a significant role in postural stabilization [8]. It was hypothesized that insufficient motor control could give rise to pain from impaired load transfer throughout the pelvic girdle [9], and that pelvic floor and abdominal muscles play an important role in the stabilization and motor control of the pelvis [8, 10]. However, there is still little evidence to support these associations. While some studies confirm the relationship between the DRA, linea alba dysfunction, and postpartum PGP [11, 12], others do not [13–15]. A recent systematic review concluded that DRA presence might be associated with decreased abdominal muscle strength and severity of low back pain and suggested further studies rigorously assessing this association [16]. A literature review investigating the relationship between perineal characteristics and PGP suggested that overactivity and increased tension of pelvic floor muscles are more common in women with PGP [17].\nThis study aimed to assess the prevalence and severity of PGP among women in Poland early postpartum and 6 weeks after delivery. Additionally, we aimed to identify factors associated with early postpartum PGP, including also DRA and pelvic floor function.","This was a prospective, observational study. Ethics approval was received from the Bioethics Committee of the Medical University of Warsaw (KB/136/2017) and the study was registered at https://www.anzctr.org.au/ under the number ACTRN 12,618,000,764,235. Data was collected between 1.12.2017 and 12.03.2020. All participants provided written informed consent prior to commencing any of the study procedures. The study was supported by the Department of Midwifery at the Centre of Postgraduate Medical Education Research Program for 2020. The STROBE checklist was followed to ensure proper reporting of this study [18].\n[SUBTITLE] Setting [SUBSECTION] St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\nSt. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\n[SUBTITLE] Participants [SUBSECTION] Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\nWomen between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\n[SUBTITLE] Procedure [SUBSECTION] The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\nThe study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\n[SUBTITLE] Measures/Variables [SUBSECTION] [SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\n[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\n[SUBTITLE] Statistical analysis [SUBSECTION] The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\nThe sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.","St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.","Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.","The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.","[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].","During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.","A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.","The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].","The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.","This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.","Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59]."],"string":"[\n \"Symptoms of pelvic girdle pain (PGP) are commonly reported to healthcare providers by pregnant or postpartum women. This musculoskeletal disorder is a form of low back pain experienced between the posterior iliac crests and the lower edge of the gluteal folds, most commonly in the vicinity of the sacroiliac joints. PGP also includes pain in the pubic symphysis, occurring in isolation or in conjunction with other pelvic joints. Patients with PGP have reduced tolerance to standing, walking, sitting, and changing positions [1]. Pregnancy-related PGP may appear as early as the first trimester of pregnancy or can be delayed up to 3 weeks postpartum [2]. PGP is associated with significantly more pain and functional limitations than lower back pain [3]. The pain often subsides after delivery, but some women continue to have persistent symptoms postpartum. Among women reporting PGP during pregnancy, 1 in 10 will suffer from it up to 11 years later [4]. This significantly impacts their quality of life [5]. In Poland, the prevalence of PGP during pregnancy was reported by 42% of women [6]. To our knowledge, there are no studies reporting on the prevalence of postpartum PGP in Poland or any other country in the central-eastern region of Europe.\\nAssessment of muscle impairments has been recommended in Clinical Practice Guidelines for PGP in the postpartum population [7]. To date, several studies investigated musculoskeletal factors that could be associated with pregnancy-related PGP, including the function of pelvic floor muscles and diastasis rectus abdominis (DRA). Coordination between lumbopelvic and abdominal muscles and fascia was suggested to play a significant role in postural stabilization [8]. It was hypothesized that insufficient motor control could give rise to pain from impaired load transfer throughout the pelvic girdle [9], and that pelvic floor and abdominal muscles play an important role in the stabilization and motor control of the pelvis [8, 10]. However, there is still little evidence to support these associations. While some studies confirm the relationship between the DRA, linea alba dysfunction, and postpartum PGP [11, 12], others do not [13–15]. A recent systematic review concluded that DRA presence might be associated with decreased abdominal muscle strength and severity of low back pain and suggested further studies rigorously assessing this association [16]. A literature review investigating the relationship between perineal characteristics and PGP suggested that overactivity and increased tension of pelvic floor muscles are more common in women with PGP [17].\\nThis study aimed to assess the prevalence and severity of PGP among women in Poland early postpartum and 6 weeks after delivery. Additionally, we aimed to identify factors associated with early postpartum PGP, including also DRA and pelvic floor function.\",\n \"This was a prospective, observational study. Ethics approval was received from the Bioethics Committee of the Medical University of Warsaw (KB/136/2017) and the study was registered at https://www.anzctr.org.au/ under the number ACTRN 12,618,000,764,235. Data was collected between 1.12.2017 and 12.03.2020. All participants provided written informed consent prior to commencing any of the study procedures. The study was supported by the Department of Midwifery at the Centre of Postgraduate Medical Education Research Program for 2020. The STROBE checklist was followed to ensure proper reporting of this study [18].\\n[SUBTITLE] Setting [SUBSECTION] St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\\nSt. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\\n[SUBTITLE] Participants [SUBSECTION] Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\\nWomen between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\\n[SUBTITLE] Procedure [SUBSECTION] The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\\nThe study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\\n[SUBTITLE] Measures/Variables [SUBSECTION] [SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\n[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\n[SUBTITLE] Statistical analysis [SUBSECTION] The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\\nThe sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\",\n \"St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\",\n \"Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\",\n \"The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\",\n \"[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\",\n \"During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\",\n \"A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\",\n \"The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\",\n \"The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\",\n \"This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\",\n \"Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Setting","Participants","Procedure","Measures/Variables","PGP prevalence and severity","Pelvic floor measurements","DRA measurements","Statistical analysis","Results","Discussion","Strengths and limitations","Implications","Conclusion"],"string":"[\n \"Background\",\n \"Methods\",\n \"Setting\",\n \"Participants\",\n \"Procedure\",\n \"Measures/Variables\",\n \"PGP prevalence and severity\",\n \"Pelvic floor measurements\",\n \"DRA measurements\",\n \"Statistical analysis\",\n \"Results\",\n \"Discussion\",\n \"Strengths and limitations\",\n \"Implications\",\n \"Conclusion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Symptoms of pelvic girdle pain (PGP) are commonly reported to healthcare providers by pregnant or postpartum women. This musculoskeletal disorder is a form of low back pain experienced between the posterior iliac crests and the lower edge of the gluteal folds, most commonly in the vicinity of the sacroiliac joints. PGP also includes pain in the pubic symphysis, occurring in isolation or in conjunction with other pelvic joints. Patients with PGP have reduced tolerance to standing, walking, sitting, and changing positions [1]. Pregnancy-related PGP may appear as early as the first trimester of pregnancy or can be delayed up to 3 weeks postpartum [2]. PGP is associated with significantly more pain and functional limitations than lower back pain [3]. The pain often subsides after delivery, but some women continue to have persistent symptoms postpartum. Among women reporting PGP during pregnancy, 1 in 10 will suffer from it up to 11 years later [4]. This significantly impacts their quality of life [5]. In Poland, the prevalence of PGP during pregnancy was reported by 42% of women [6]. To our knowledge, there are no studies reporting on the prevalence of postpartum PGP in Poland or any other country in the central-eastern region of Europe.\nAssessment of muscle impairments has been recommended in Clinical Practice Guidelines for PGP in the postpartum population [7]. To date, several studies investigated musculoskeletal factors that could be associated with pregnancy-related PGP, including the function of pelvic floor muscles and diastasis rectus abdominis (DRA). Coordination between lumbopelvic and abdominal muscles and fascia was suggested to play a significant role in postural stabilization [8]. It was hypothesized that insufficient motor control could give rise to pain from impaired load transfer throughout the pelvic girdle [9], and that pelvic floor and abdominal muscles play an important role in the stabilization and motor control of the pelvis [8, 10]. However, there is still little evidence to support these associations. While some studies confirm the relationship between the DRA, linea alba dysfunction, and postpartum PGP [11, 12], others do not [13–15]. A recent systematic review concluded that DRA presence might be associated with decreased abdominal muscle strength and severity of low back pain and suggested further studies rigorously assessing this association [16]. A literature review investigating the relationship between perineal characteristics and PGP suggested that overactivity and increased tension of pelvic floor muscles are more common in women with PGP [17].\nThis study aimed to assess the prevalence and severity of PGP among women in Poland early postpartum and 6 weeks after delivery. Additionally, we aimed to identify factors associated with early postpartum PGP, including also DRA and pelvic floor function.","This was a prospective, observational study. Ethics approval was received from the Bioethics Committee of the Medical University of Warsaw (KB/136/2017) and the study was registered at https://www.anzctr.org.au/ under the number ACTRN 12,618,000,764,235. Data was collected between 1.12.2017 and 12.03.2020. All participants provided written informed consent prior to commencing any of the study procedures. The study was supported by the Department of Midwifery at the Centre of Postgraduate Medical Education Research Program for 2020. The STROBE checklist was followed to ensure proper reporting of this study [18].\n[SUBTITLE] Setting [SUBSECTION] St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\nSt. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\n[SUBTITLE] Participants [SUBSECTION] Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\nWomen between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\n[SUBTITLE] Procedure [SUBSECTION] The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\nThe study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\n[SUBTITLE] Measures/Variables [SUBSECTION] [SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\n[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\n[SUBTITLE] Statistical analysis [SUBSECTION] The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\nThe sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.","St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.","Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.","The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.","[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].","During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\n\nFig. 1Pain maps used for assessing the presence of PGP\n\nPain maps used for assessing the presence of PGP\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.","A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.","The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\n\nFig. 2DRA and linea alba assessment\n\nDRA and linea alba assessment\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].","The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$n=\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\end{document}\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.","A total of 415 women were invited to participate, 4 of whom did not consent to join the study. Thus, 411 women were included in phase 1 (shortly postpartum). Table 1. presents the characteristics of the participants. In phase 2 (“follow-up”), 268 women replied to the text message (65.2% of the initial group). No statistically significant differences were found between those who responded to the message and those who did not in terms of variables assessed shortly postpartum (in phase 1): BMI, age, presence of PGP postpartum, parity, DRA severity, and ability to activate pelvic floor muscles.\n\nTable 1Characteristics of the study group, n = 411All participantsn = 411With PGP n = 37Without PGP n = 374\nAge\n31.17 ± 4.0132.35 ± 4.3331.06 ± 3.96Height [cm],167.46 ± 5.69167.00 ± 6.64167.51 ± 5.60\nBMI before pregnancy\n22.14 ± 3.5722.46 ± 3.2022.10 ± 3.60\nBody mass gain in pregnancy [kg]\n14.30 ± 4.9414.92 ± 4.2814.24 ± 4.99Education, n (%) †vocational education2 (0.5)11secondary education43 (10.5)242university education366 (89)35331\nParity, n (%)\n1.72 ± 0.961.95 ± 1.101.70 ± 0.941209 (50.80)2143 (34.80)337 (9.00)412 (3.00)58 (2.00)61 (0.20)71 (0.20)\nMode of the last delivery, n (%)\nvaginal387 (94.20)37 (100)350 (93.60)cesarean ‡18 (4.30)018 (4.80)vacuum extractor ‡6 (1.50)06 (1.60)\nPerineal injury during recent delivery, n (%)\nnone162 (39.50)18 (48.60)144 (38.50)1st grade138 (33.50)9 (24.30)129 (34.50)episiotomy111 (27)10 (27)101 (27)\nAnesthesia during last delivery, n (%)\nnone238 (57.91)24 (64.86)214 (57.22)epidural154 (37.47)12 (32.43)142 (37.97)spinal19 (4.62)1 (2.70)18 (4.81)\nInfant body mass ≥ 4000 g\nyes336 (81.75)6 (16.22)330 (88.24)no75 (18.25)31 (83.78)44 (11.76)† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\n\nCharacteristics of the study group, n = 411\n† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\nAmong the patients in phase 1 of the study (shortly postpartum), 47.9% (n = 197) reported PGP symptoms during pregnancy. In phase 1, PGP was diagnosed in 9.0% (n = 37) of women at the early postpartum stage. In phase 2 (6 weeks postpartum), PGP was reported by 15.7% of women (n = 42). Table 2 presents the prevalence and type of PGP across different time points, and Table 3 shows the severity of pain and functional disturbances in both phases of the study.\n\nTable 2Prevalence and types of PGP across the time points.TimepointDuring pregnancyEarly postpartum6 weeks postpartum\nAssessment form\n\nself-reported, retrospective\n\nclinical assessment\n\nself-reported\n\nN\n\n%\n\nN\n\n%\n\nN\n\n%\n\nPGP\n197/41147.9337/411942/26815.70\nPGP\n\ntype\n\nPosterior Pelvic Pain\n86/19743.709/3724.3016/4238.10\nUnilateral pain\n17/1978.602/375.402/424.80\nSymphyseal pain\n44/19722.309/3724.3012/4228.60\nPelvic Girdle Syndrome\n47/19723.9016/3743.2011/4226.20\nUnilateral pain + symphyseal pain\n3/1971.501/372.701/422.40\n\nPrevalence and types of PGP across the time points.\n\nPGP\n\n\ntype\n\n\nTable 3The severity of pain and functional limitations at phase 1 and 2 of the studyTimepointEarlyPostpartum6 weekspostpartum\nMean (SD)\n\n(min-max)\n\nN\n\nMean (SD)\n\n(min-max)\n\nN\n\nNRS\n5.34 (1.74)(3–9)37/374.63 (2.04)(1–8)24/42\nPGQ [%]\n48.87 (16.40)(23.61–82.61)32/3725.80 (14.90)(6.67–56.94)13/42The values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\n\nThe severity of pain and functional limitations at phase 1 and 2 of the study\n5.34 (1.74)\n(3–9)\n4.63 (2.04)\n(1–8)\n48.87 (16.40)\n(23.61–82.61)\n25.80 (14.90)\n(6.67–56.94)\nThe values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\nThe univariable analyses showed a higher likelihood of PGP shortly postpartum in women who declared PGP during pregnancy and among women with doming at the abdominal midline in the projection of linea alba (Table 4.).\n\nTable 4Univariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)PGP (+)n = 37PGP (-)N = 374OR (95% CI)p-value\nPGP during pregnancy (yes)\n\nn (%)\n34 (91.90)163 (43.60)14.67 (4.43–48.61)\n< 0.01\n\nAge\n\nmean (SD)\n32.35(4.33)31.06 (3.96)1.08 (0.99–1.18)0.06\nBody mass gain during pregnancy\n\nmean (SD)\n32.68 (4.60)31.59 (4.37)1.03 (0.96–1.10)0.42\nBMI before pregnancy\n\nmean (SD)\n22.46 (3.20)22.1 (3.60)1.03 (0.94–1.12)0.56\nNumber of previous deliveries mean (SD)\n1.95 (1.10)1.7 (0.94)1.27 (0.93–1.71)0.14\nUrinary incontinence during pregnancy or before\n\n(yes) n (%)\n15 (40.50)205 (54.80)0.56 (0.28–1.12)0.10\nAbdominal midline doming (yes), n (%)\n20 (54.10)134 (36.40)2.05 (1.04–4.06)\n0.04\n\nDR severity\n\nn (%)\n\nnone, IRD < 2\n9 (24.30)149 (40.50)1.27 (0.95–1.69)0.11mild,\nIRD 2;< 3\n8 (21.60)72 (19.60)\nmoderate, IRD 3;<4\n12 (32.40)75 (20.40)\nsevere, IRD > 4\n8 (21.60)72 (19.60)\nOxford scale\n\nMean (SD)\n2.27(0.65)2.28 (0.90)0.99 (0.67–1.45)0.96\nReissing scale\n\nmean (SD)\n0.32 (1.11)-0.53 (0.96)1.25 (0.88–1.77)0.21\nCorrect activation of pelvic floor (yes), n (%)\n13 (35.10)134 (35.80)1.36 (0.69–2.70)0.37\n\nUnivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\n\nPGP during pregnancy (yes)\n\n\nn (%)\n\n\nAge\n\n\nmean (SD)\n\n\nBody mass gain during pregnancy\n\n\nmean (SD)\n\n\nBMI before pregnancy\n\n\nmean (SD)\n\n\nUrinary incontinence during pregnancy or before\n\n\n(yes) n (%)\n\n\nDR severity\n\n\nn (%)\n\nmild,\n\nIRD 2;< 3\n\n\nOxford scale\n\n\nMean (SD)\n\n\nReissing scale\n\n\nmean (SD)\n\nBased on the obtained prevalence of postpartum PGP, we could include up to 4 variables into the multivariable model [38]. The final model, in which we obtained statistically significant results for all included items, consisted of 3 variables. This multivariable regression analysis showed that the odds of having PGP shortly postpartum were higher in women with increased age (10% higher likelihood with every year of age), declaring PGP during pregnancy, and with higher values of Reissing scale. To enhance the interpretability, we then analyzed the Reissing scale in the following categories: hypotonus (range − 3 to -1) and increased muscle tone (range 1 to 3), with the reference value being normotonus (0). However, after this procedure, the Reissing scale became not statistically significant (Table 5). The calculated Nagelkerke R-square was 0.2.\n\nTable 5Multivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)OR (95% CI)p-value\nPresence of PGP during pregnancy (yes)\n14.83 (4.340-48.721)\n< 0.0001\n\nAge\n1.12 (1.009–1.214)\n0.032\n\nReissing scale\n\nReissing scale – not grouped\n1.43 (1.003–2.046)\n0.048\n\nReissing scale when grouped\n\nnormal tone (0)\nreferencereference\ndecreased tone \n\n(range − 3 to -1)\n0.53 (0.227–1.220)0.134\nincreased tone \n\n(range 1 to 3)\n0.37 (0.594–4.078)0.368\n\nMultivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\n\ndecreased tone \n\n\n(range − 3 to -1)\n\n\nincreased tone \n\n\n(range 1 to 3)\n","Our study showed that nearly 10% of women were diagnosed with PGP during the first days postpartum, and almost 16% reported similar symptoms 6 weeks later. The mean pain intensity and functional limitations within the first days postpartum were moderate, with values corresponding to mild/low 6 weeks postpartum. The likelihood of experiencing PGP shortly after delivery increased with age and reporting PGP during pregnancy. The doming of the abdominal midline was significantly associated with PGP shortly postpartum only in the univariable analysis. The remaining variables related to diastasis recti or pelvic floor function were not associated with PGP shortly after delivery (24–72 h postpartum).\nThe reported prevalence of postpartum PGP around 12 weeks after delivery ranges from 3.4 to 43.0% [39–46] with the majority of studies [41–45] showing a higher prevalence than demonstrated in this research. This variation may be due to several reasons. Firstly, diverse diagnoses and terminology were used in the mentioned studies, which could lead to discrepancies in prevalence. In our research, postpartum PGP was defined as a pain that persisted postpartum or occurred within the first weeks after delivery [2]. However, in the study of Stomp van den Berg et al. [44] 25% of the 234 women who had PGP at 12 weeks postpartum had no PGP between 0 and 6 weeks after delivery. Secondly, cultural and ethnic factors can play a role in the processes related to pain perception [47]. Although PGP is prevalent worldwide, it is not recognized by health care systems in some countries. Our previous study has shown that PGP during pregnancy was more common in Norwegian than Polish women [6]. In Norway, PGP is one of the most common causes of sick leave among pregnant women [48]. In Poland, PGP is not commonly recognized, and the term ‘pelvic girdle pain’ is not widely used within health care services. Lower social awareness about this condition could lead to lower reporting. The possible role of ethnicity was noticed in another PGP study [49] indicating a more detailed investigation encompassing cultural and ethnic influences associated with PGP is needed.\nWe could observe similar discrepancies when analyzing the values related to pain intensity and functional limitations, possibly related to the same reasons as those mentioned above. For instance, in the study by Mukkanavar et al. [41] among Indian postpartum women, as many as 84.5% participants with PGP between the 3rd and 18th week after delivery rated their symptoms as greater than 60 mm on the VAS scale. Stomp van de Berg [44] reported median pain intensity 6 weeks postpartum at 4.3 of NRS scale. In the study of Dunn et al. [42] the mean pain intensity values measured on VAS scale were between 22.5 and 55, depending on the location of PGP and co-existing dysfunctions. Our results seem to be in line with those of Sakamoto et al. [50] who also measured functional limitations with PGQ. In the second day postpartum the mean values were oscillating around 47% (95%CI 40–54), while 4 weeks after − 19% (95%CI 12–25).\nOur results showing higher PGP prevalence 6 weeks postpartum when compared to early postpartum period may seem contradictory to previous reports [37, 51]. However, it has to be noted that the cited studies followed women experiencing PGP already during pregnancy. The occurrence of pregnancy-related PGP may be delayed up to the first weeks postpartum [2]. By following all women (with and without pain), our study could capture those individuals that developed pain after the initial examination, 24-72 h postpartum. Additionally, early postpartum period is associated with more bed rest when compared to 6 weeks postpartum when PGP symptoms could be more noticeable and bothersome.\nWhen it comes to factors associated with postpartum PGP, our results are in line with previous reports. A recent systematic review by Wiezer et al. [52] confirmed PGP during pregnancy as a risk factor for persistent postpartum pain. These findings suggest that asking women whether they have experienced PGP during pregnancy may help identify women at risk of persistent postpartum pain. The association between postpartum PGP and age has also been previously shown. Gausel et al. [46] reported age 30 and above as the risk factors for persistent postpartum pain. In European countries, primiparous women are becoming older. In Poland, the mean age of women having their first baby in 2019 was 27.4 and, although constantly increasing, is still one of the lowest in Europe [53]. Increasing maternal age may have several consequences. In accordance with previously mentioned studies, our results indicate that pregnant women who deliver past a certain age should receive special physiotherapy care.\nIn our study, the doming of the abdominal wall in the projection of linea alba was a statistically significant factor only in univariable analysis and there were no associations between the DRA severity (size of IRD) and the presence of PGP shortly postpartum. This is different when compared to our recently published matched-case control studies [54, 55]. However, in mentioned reports participants were matched according to age and parity, mode of delivery and time postpartum. This may suggest that although DRA features and postpartum PGP may co-exist, those associations are not straightforward and there are possibly other factors that may mediate this relationship.\nOur study did not reveal any associations between pelvic floor function and PGP shortly postpartum, despite previous reports [17, 55]. This may be due to the timing of phase 1 of the study when the measurements were taken - some differences in the pelvic floor function may be too subtle to be detected using screening palpation examination in the early postpartum period. Our other hypothesis is that there are no differences in the pelvic floor muscle function between women with and without postpartum PGP shortly after delivery. They may be more visible with time, while pain persists and the adaptive changes in the pelvic floor occur, which could be supported by our other PGP study [55].\n[SUBTITLE] Strengths and limitations [SUBSECTION] This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\nThis was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\n[SUBTITLE] Implications [SUBSECTION] Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].\nObtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].","This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.","Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].","The findings presented in this study suggest that every tenth Polish woman may experience PGP during the first days postpartum and every sixth can report similar symptoms 6 weeks later. The pain intensity and functional limitations tend to subside over time: from moderate pain intensity and functional limitation shortly postpartum to mild/low 6 weeks later. Nevertheless, postpartum PGP should not be ignored, especially in the context of the observed continued increase in chronic pain syndromes and their associated consequences. Older age, PGP during pregnancy, and doming of the abdominal midline at the level of linea alba were associated with the experience of PGP within the first days postpartum. Our study showed no association between pelvic floor function and PGP shortly postpartum. However, this may be due to the chosen methodology (assessment shortly postpartum)."],"string":"[\n \"Symptoms of pelvic girdle pain (PGP) are commonly reported to healthcare providers by pregnant or postpartum women. This musculoskeletal disorder is a form of low back pain experienced between the posterior iliac crests and the lower edge of the gluteal folds, most commonly in the vicinity of the sacroiliac joints. PGP also includes pain in the pubic symphysis, occurring in isolation or in conjunction with other pelvic joints. Patients with PGP have reduced tolerance to standing, walking, sitting, and changing positions [1]. Pregnancy-related PGP may appear as early as the first trimester of pregnancy or can be delayed up to 3 weeks postpartum [2]. PGP is associated with significantly more pain and functional limitations than lower back pain [3]. The pain often subsides after delivery, but some women continue to have persistent symptoms postpartum. Among women reporting PGP during pregnancy, 1 in 10 will suffer from it up to 11 years later [4]. This significantly impacts their quality of life [5]. In Poland, the prevalence of PGP during pregnancy was reported by 42% of women [6]. To our knowledge, there are no studies reporting on the prevalence of postpartum PGP in Poland or any other country in the central-eastern region of Europe.\\nAssessment of muscle impairments has been recommended in Clinical Practice Guidelines for PGP in the postpartum population [7]. To date, several studies investigated musculoskeletal factors that could be associated with pregnancy-related PGP, including the function of pelvic floor muscles and diastasis rectus abdominis (DRA). Coordination between lumbopelvic and abdominal muscles and fascia was suggested to play a significant role in postural stabilization [8]. It was hypothesized that insufficient motor control could give rise to pain from impaired load transfer throughout the pelvic girdle [9], and that pelvic floor and abdominal muscles play an important role in the stabilization and motor control of the pelvis [8, 10]. However, there is still little evidence to support these associations. While some studies confirm the relationship between the DRA, linea alba dysfunction, and postpartum PGP [11, 12], others do not [13–15]. A recent systematic review concluded that DRA presence might be associated with decreased abdominal muscle strength and severity of low back pain and suggested further studies rigorously assessing this association [16]. A literature review investigating the relationship between perineal characteristics and PGP suggested that overactivity and increased tension of pelvic floor muscles are more common in women with PGP [17].\\nThis study aimed to assess the prevalence and severity of PGP among women in Poland early postpartum and 6 weeks after delivery. Additionally, we aimed to identify factors associated with early postpartum PGP, including also DRA and pelvic floor function.\",\n \"This was a prospective, observational study. Ethics approval was received from the Bioethics Committee of the Medical University of Warsaw (KB/136/2017) and the study was registered at https://www.anzctr.org.au/ under the number ACTRN 12,618,000,764,235. Data was collected between 1.12.2017 and 12.03.2020. All participants provided written informed consent prior to commencing any of the study procedures. The study was supported by the Department of Midwifery at the Centre of Postgraduate Medical Education Research Program for 2020. The STROBE checklist was followed to ensure proper reporting of this study [18].\\n[SUBTITLE] Setting [SUBSECTION] St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\\nSt. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\\n[SUBTITLE] Participants [SUBSECTION] Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\\nWomen between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\\n[SUBTITLE] Procedure [SUBSECTION] The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\\nThe study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\\n[SUBTITLE] Measures/Variables [SUBSECTION] [SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\n[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\n[SUBTITLE] Statistical analysis [SUBSECTION] The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\\nThe sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\",\n \"St. Sophia Specialist Hospital in Warsaw, Poland served as a recruitment site for this study. It is a tertiary, publicly funded hospital with over 6500 births annually. A free consultation with a pelvic health physiotherapist is part of standard care for every woman after delivery at this hospital. The participants were recruited from among the women attending the consultation.\",\n \"Women between 18 and 45 years old who attended free physiotherapy consultation 24–72 h postpartum were invited to participate in the study. The main exclusion criteria were additional comorbidities potentially causing PGP-like symptoms (rheumatoid arthritis, ankylosing spondylitis, Scheuermann disease, Ehlers-Danlos syndrome, spinal surgeries, nerve root compression, spondylolisthesis), contraindications for pelvic floor examination (puerperal genital hematoma, diffuse perineal edema, perineal wound dehiscence, bladder catheterization) and severe postpartum complications (internal bleeding, femoral artery embolism, pelvic fracture). To limit the sampling bias and ensure that patients were randomly included in the study, every third participant of the postpartum physiotherapy consultation was invited, and recruitment for the study took place every third day.\",\n \"The study was conducted in two phases. Phase 1 was carried out at the hospital. Participants who met the inclusion criteria and gave informed consent were included in the study. Age, education, parity (defined as previous deliveries > 24 weeks gestation), delivery type (vaginal, forceps/vacuum extractor, cesarean), infant body mass (< 4000 g/4000 g or more), height, body mass gain during pregnancy, the use of anesthesia during delivery were recorded from the patient medical record and confirmed via self-report. The women were asked if they had experienced PGP during the last pregnancy (yes/no and if yes - in what location) and urinary incontinence during or before the pregnancy (yes/no). The presence and severity of PGP were assessed by the physiotherapist. Additionally, each woman received an examination of the pelvic floor and abdominal muscles. All examinations were performed by the same registered pelvic health physiotherapist who had completed advanced training in urogynaecology and was certified by the Polish Urogynecological Society to digitally examine the pelvic floor muscles.\\nIn phase 2 (“follow-up”), all women who participated in phase 1 were contacted via text message 6 weeks postpartum. In the message, they were asked if they experienced PGP. In case of no response within 48 h, another message was sent. No response to the second message meant a loss to follow-up.\",\n \"[SUBTITLE] PGP prevalence and severity [SUBSECTION] During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\nDuring the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\\n[SUBTITLE] Pelvic floor measurements [SUBSECTION] A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\nA pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\\n[SUBTITLE] DRA measurements [SUBSECTION] The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\\nThe width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\",\n \"During the examination by the physiotherapist in phase 1 (shortly postpartum), every patient was asked the question: “Do you have pelvic girdle pain in the places marked in the figure (Fig. 1), which aggravates during activities such as standing up, walking, or rolling from side to side?“ The participants reported their pain by indicating its location on a body chart. This was confirmed by pointing to the site of the pain in their body. The physiotherapist then carried out a further clinical examination to confirm the presence of PGP. For PGP classification, existing guidelines and previous reports were used [1, 3, 4]. Firstly, the lumbar spine examination was performed (flexion/extension movements, lateral rotations, lateral bends, Laseque test) to exclude lumbar causes of pain in the pelvic girdle region. This was followed by the tests dedicated to the pelvic girdle: Posterior Pelvic Pain Provocation (P4) test, distraction test, compression test, palpation of the pubic symphysis, modified Trendelenburg test, active straight leg raise test (ASLR). At least two tests had to be positive for PGP to be confirmed. For the ASLR test, the scores on both sides were added, and the total score ranged from 0 to 10. An ASLR total score of 4 and above was considered positive.\\n\\nFig. 1Pain maps used for assessing the presence of PGP\\n\\nPain maps used for assessing the presence of PGP\\nPatients with confirmed PGP rated their mean pain intensity during a day on a Numerical Rating Scale (NRS) from 0 to 10, where 0 meant no pain and 10 meant the worst pain imaginable [19]. We designated the categorical cut-off points for the NRS as mild (1–4), moderate (5–6), and severe (7 to 10) [20]. Functional limitations were assessed with the Polish version of The Pelvic Girdle Questionnaire (PGQ) [21]. Values 0–28 were interpreted as low, 28–62 as moderate, and > 62 as high. [22]\\nAssessment of PGP prevalence 6 weeks postpartum was carried out via text message. To increase the certainty of our results, we have provided the figure with marked pain locations, and also a short description of PGP-related symptoms. The women were asked: “Are you currently experiencing pelvic/sacrum/coccyx/pubic symphysis pain (Fig. 1) with or without a feeling of instability in these areas, arising or worsening with changes in position or movement?”. The message was accompanied by an image with marked pain locations (Fig. 1). Women were classified as having pelvic girdle syndrome if they indicated 1st and 3rd location. If the response was positive, the patient was asked to rate the pain using NRS and functional limitations by filling out and returning the PGQ via email or MMS.\",\n \"A pelvic floor assessment was performed using the palpation examination. We decided this would be the only appropriate method to be used in the early postpartum period as it is fast, painless, non-invasive, and requires no additional equipment. All patients were assessed in a crook lying position. The PERFECT scheme [23] with the 6-point OXFORD scale (0–5) was used to evaluate a maximal voluntary contraction (MVC) of the pelvic floor. The OXFORD scale is a reliable measure of the MVC with acceptable intra-observer and test-retest reliability [24]. The existing research showed that the palpation assessment using the OXFORD scale was consistent with the ultrasound examination results [25].\\nA seven-point scale proposed by Reising et al. was used to assess muscle tone with values ranging from − 3 (very hypotonic muscles) to + 3 (very hypertonic muscles) with 0 stating normal tone. This scale has been studied for its reliability showing fair-to-moderate interrater reliability with correlation coefficients of 0.2–0.5 [26]. Weak-to-fair associations between the Reissing scale, dynamometry, and ultrasound imaging with correlation coefficients of 0.2–0.4 have been previously shown [27].\\nAdditionally, possible activation of synergistic muscles (gluteal muscles, adductors, and abdomen), and the ability to activate the pelvic floor muscles without breath-holding were observed during the pelvic floor examination. That was done in order to determine the correctness of pelvic floor activation: isolated pelvic floor contraction without breath-holding was considered correct.\",\n \"The width of rectus abdominus muscle bellies (inter-recti distance, IRD) was determined by palpation using the procedure reported in previous studies [28]. The women were asked to lie in a standard supine position with their knees bent. Then, they were asked to perform the abdominal curl-up by raising the head and upper torso until the shoulder blades left the examination bed (Fig. 2). Measurements were taken at the navel level and 4.5 cm above and below it [29, 30]. As done in previous studies, the point of largest width was selected for the analysis [13–15]. Mota et al. showed good intra-rater reliability (Kw>0.7) in terms of palpation measurements of IRD [28] and Benjamin et al. showed moderate to very good correlation of IRD palpation with ultrasound (r = 0.75–0.98) [31]. According to other studies implementing this method [13, 15, 29, 30] DRA was considered when the IRD value was ≥ 2 fingerbreadths. The participants in this study were divided into four categories depending on the largest palpation measurement (number of fingers) in one of three locations: no DRA (IRD < 2 fingerbreadths), mild DRA (IRD 2;< 3), moderate DR (IRD 3;<4), severe DRA (IRD > 4) [13].\\n\\nFig. 2DRA and linea alba assessment\\n\\nDRA and linea alba assessment\\nIt has been suggested that the integrity of linea alba may influence the capacity to stabilize the pelvis and the lumbar spine, and the ability of the linea alba to transmit forces across the midline may have a more significant impact on function than the magnitude of the IRD [32]. Therefore, in the context of the assessment of DRA and linea alba dysfunction, not only IRD but also the stiffness and distortion/bulging at the level of linea alba could play an important role [32, 33]. To investigate any possible association with PGP, we have adopted a simplified method of assessing the doming of the abdominal midline. It was defined as “abdominal midline doming” (yes/no). During the curl-up test, the physiotherapist observed whether the abdominal midline bulged. Although this method does not allow to determine which structures are bulging (e.g., linea alba or “just” subcutaneous tissue), it may give us a simplified estimation of pressure management in the abdominal cavity. A similar assessment has recently been used in another research [34].\",\n \"The sample size was determined based on a priori calculations, based on the reports of Wu et al., in which the incidence of postpartum low back pain and /or PGP is estimated at 25% [35]. The following formula was used:\\\\documentclass[12pt]{minimal}\\n\\t\\t\\t\\t\\\\usepackage{amsmath}\\n\\t\\t\\t\\t\\\\usepackage{wasysym} \\n\\t\\t\\t\\t\\\\usepackage{amsfonts} \\n\\t\\t\\t\\t\\\\usepackage{amssymb} \\n\\t\\t\\t\\t\\\\usepackage{amsbsy}\\n\\t\\t\\t\\t\\\\usepackage{mathrsfs}\\n\\t\\t\\t\\t\\\\usepackage{upgreek}\\n\\t\\t\\t\\t\\\\setlength{\\\\oddsidemargin}{-69pt}\\n\\t\\t\\t\\t\\\\begin{document}$$n=\\\\frac{{Z}^{2}P(1-P)}{{d}^{2}},$$\\\\end{document}\\nwhere n is the sample size, the Z value (corresponding to the significance level p 0.05) is 1.96. P value is the expected occurrence of pain, and the d value is 0.05 (recommended value for an expected prevalence between 10 and 90% [36]). Based on this calculation, the minimum sample size was 288. A high drop-out rate was also expected due to the specificity of the postpartum period, as observed by other authors [37]. To secure the minimal number of participants for phase 2 of the study (“follow-up”), the sample size was increased to a minimum of 400.\\nFor continuous variables, mean and SD were calculated. Categorical data are presented as numbers and percentages. The prevalence of PGP was calculated by dividing the number of women classified with PGP by the total number of women who participated in the study. To assess the possible differences between women who responded 6 weeks postpartum and those who were lost to follow-up, Student t-test, the Mann-Whitney U and the the χ2 test were used depending on the type and normality of data.\\nUnivariable logistic regression analyses were fitted to test individual factors for association with PGP shortly postpartum (phase 1). Multivariable logistic models were used to identify associations with more than one factor included in the model. The selection of factors to include in the multivariable models was informed by the univariable results. Variables with p-values less than or equal to 0.1 were included in the multivariable regression model. The number of possible investigated factors in the multivariable model was calculated assuming 10 participants per potential associated factor with division by the obtained prevalence rate [38]. The best subset of explanatory variables was selected manually by excluding the variables with the smallest contribution to the model. The predictive power of the model was calculated by Nagelkerke R-square (R²). Missing data were not included in the analysis. Alpha was set at 0.05. Statistical analyses were performed using PQStat ver. 1.8.2.166.\",\n \"A total of 415 women were invited to participate, 4 of whom did not consent to join the study. Thus, 411 women were included in phase 1 (shortly postpartum). Table 1. presents the characteristics of the participants. In phase 2 (“follow-up”), 268 women replied to the text message (65.2% of the initial group). No statistically significant differences were found between those who responded to the message and those who did not in terms of variables assessed shortly postpartum (in phase 1): BMI, age, presence of PGP postpartum, parity, DRA severity, and ability to activate pelvic floor muscles.\\n\\nTable 1Characteristics of the study group, n = 411All participantsn = 411With PGP n = 37Without PGP n = 374\\nAge\\n31.17 ± 4.0132.35 ± 4.3331.06 ± 3.96Height [cm],167.46 ± 5.69167.00 ± 6.64167.51 ± 5.60\\nBMI before pregnancy\\n22.14 ± 3.5722.46 ± 3.2022.10 ± 3.60\\nBody mass gain in pregnancy [kg]\\n14.30 ± 4.9414.92 ± 4.2814.24 ± 4.99Education, n (%) †vocational education2 (0.5)11secondary education43 (10.5)242university education366 (89)35331\\nParity, n (%)\\n1.72 ± 0.961.95 ± 1.101.70 ± 0.941209 (50.80)2143 (34.80)337 (9.00)412 (3.00)58 (2.00)61 (0.20)71 (0.20)\\nMode of the last delivery, n (%)\\nvaginal387 (94.20)37 (100)350 (93.60)cesarean ‡18 (4.30)018 (4.80)vacuum extractor ‡6 (1.50)06 (1.60)\\nPerineal injury during recent delivery, n (%)\\nnone162 (39.50)18 (48.60)144 (38.50)1st grade138 (33.50)9 (24.30)129 (34.50)episiotomy111 (27)10 (27)101 (27)\\nAnesthesia during last delivery, n (%)\\nnone238 (57.91)24 (64.86)214 (57.22)epidural154 (37.47)12 (32.43)142 (37.97)spinal19 (4.62)1 (2.70)18 (4.81)\\nInfant body mass ≥ 4000 g\\nyes336 (81.75)6 (16.22)330 (88.24)no75 (18.25)31 (83.78)44 (11.76)† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\\n\\nCharacteristics of the study group, n = 411\\n† Participants with vocational education were excluded (from this calculations) because of the low number of records; ‡ no statistical analysis due to low number of observations; χ² Chi2 tet value; U Manna Whitney test value\\nAmong the patients in phase 1 of the study (shortly postpartum), 47.9% (n = 197) reported PGP symptoms during pregnancy. In phase 1, PGP was diagnosed in 9.0% (n = 37) of women at the early postpartum stage. In phase 2 (6 weeks postpartum), PGP was reported by 15.7% of women (n = 42). Table 2 presents the prevalence and type of PGP across different time points, and Table 3 shows the severity of pain and functional disturbances in both phases of the study.\\n\\nTable 2Prevalence and types of PGP across the time points.TimepointDuring pregnancyEarly postpartum6 weeks postpartum\\nAssessment form\\n\\nself-reported, retrospective\\n\\nclinical assessment\\n\\nself-reported\\n\\nN\\n\\n%\\n\\nN\\n\\n%\\n\\nN\\n\\n%\\n\\nPGP\\n197/41147.9337/411942/26815.70\\nPGP\\n\\ntype\\n\\nPosterior Pelvic Pain\\n86/19743.709/3724.3016/4238.10\\nUnilateral pain\\n17/1978.602/375.402/424.80\\nSymphyseal pain\\n44/19722.309/3724.3012/4228.60\\nPelvic Girdle Syndrome\\n47/19723.9016/3743.2011/4226.20\\nUnilateral pain + symphyseal pain\\n3/1971.501/372.701/422.40\\n\\nPrevalence and types of PGP across the time points.\\n\\nPGP\\n\\n\\ntype\\n\\n\\nTable 3The severity of pain and functional limitations at phase 1 and 2 of the studyTimepointEarlyPostpartum6 weekspostpartum\\nMean (SD)\\n\\n(min-max)\\n\\nN\\n\\nMean (SD)\\n\\n(min-max)\\n\\nN\\n\\nNRS\\n5.34 (1.74)(3–9)37/374.63 (2.04)(1–8)24/42\\nPGQ [%]\\n48.87 (16.40)(23.61–82.61)32/3725.80 (14.90)(6.67–56.94)13/42The values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\\n\\nThe severity of pain and functional limitations at phase 1 and 2 of the study\\n5.34 (1.74)\\n(3–9)\\n4.63 (2.04)\\n(1–8)\\n48.87 (16.40)\\n(23.61–82.61)\\n25.80 (14.90)\\n(6.67–56.94)\\nThe values of the PGQ questionnaire from the early postpartum stage were missing in five cases – they were left blank or only partially completed. Six weeks after delivery, the pain intensity value on the NRS was reported by 24 women, and the PGQ questionnaire was completed only by 13\\nThe univariable analyses showed a higher likelihood of PGP shortly postpartum in women who declared PGP during pregnancy and among women with doming at the abdominal midline in the projection of linea alba (Table 4.).\\n\\nTable 4Univariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)PGP (+)n = 37PGP (-)N = 374OR (95% CI)p-value\\nPGP during pregnancy (yes)\\n\\nn (%)\\n34 (91.90)163 (43.60)14.67 (4.43–48.61)\\n< 0.01\\n\\nAge\\n\\nmean (SD)\\n32.35(4.33)31.06 (3.96)1.08 (0.99–1.18)0.06\\nBody mass gain during pregnancy\\n\\nmean (SD)\\n32.68 (4.60)31.59 (4.37)1.03 (0.96–1.10)0.42\\nBMI before pregnancy\\n\\nmean (SD)\\n22.46 (3.20)22.1 (3.60)1.03 (0.94–1.12)0.56\\nNumber of previous deliveries mean (SD)\\n1.95 (1.10)1.7 (0.94)1.27 (0.93–1.71)0.14\\nUrinary incontinence during pregnancy or before\\n\\n(yes) n (%)\\n15 (40.50)205 (54.80)0.56 (0.28–1.12)0.10\\nAbdominal midline doming (yes), n (%)\\n20 (54.10)134 (36.40)2.05 (1.04–4.06)\\n0.04\\n\\nDR severity\\n\\nn (%)\\n\\nnone, IRD < 2\\n9 (24.30)149 (40.50)1.27 (0.95–1.69)0.11mild,\\nIRD 2;< 3\\n8 (21.60)72 (19.60)\\nmoderate, IRD 3;<4\\n12 (32.40)75 (20.40)\\nsevere, IRD > 4\\n8 (21.60)72 (19.60)\\nOxford scale\\n\\nMean (SD)\\n2.27(0.65)2.28 (0.90)0.99 (0.67–1.45)0.96\\nReissing scale\\n\\nmean (SD)\\n0.32 (1.11)-0.53 (0.96)1.25 (0.88–1.77)0.21\\nCorrect activation of pelvic floor (yes), n (%)\\n13 (35.10)134 (35.80)1.36 (0.69–2.70)0.37\\n\\nUnivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\\n\\nPGP during pregnancy (yes)\\n\\n\\nn (%)\\n\\n\\nAge\\n\\n\\nmean (SD)\\n\\n\\nBody mass gain during pregnancy\\n\\n\\nmean (SD)\\n\\n\\nBMI before pregnancy\\n\\n\\nmean (SD)\\n\\n\\nUrinary incontinence during pregnancy or before\\n\\n\\n(yes) n (%)\\n\\n\\nDR severity\\n\\n\\nn (%)\\n\\nmild,\\n\\nIRD 2;< 3\\n\\n\\nOxford scale\\n\\n\\nMean (SD)\\n\\n\\nReissing scale\\n\\n\\nmean (SD)\\n\\nBased on the obtained prevalence of postpartum PGP, we could include up to 4 variables into the multivariable model [38]. The final model, in which we obtained statistically significant results for all included items, consisted of 3 variables. This multivariable regression analysis showed that the odds of having PGP shortly postpartum were higher in women with increased age (10% higher likelihood with every year of age), declaring PGP during pregnancy, and with higher values of Reissing scale. To enhance the interpretability, we then analyzed the Reissing scale in the following categories: hypotonus (range − 3 to -1) and increased muscle tone (range 1 to 3), with the reference value being normotonus (0). However, after this procedure, the Reissing scale became not statistically significant (Table 5). The calculated Nagelkerke R-square was 0.2.\\n\\nTable 5Multivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)OR (95% CI)p-value\\nPresence of PGP during pregnancy (yes)\\n14.83 (4.340-48.721)\\n< 0.0001\\n\\nAge\\n1.12 (1.009–1.214)\\n0.032\\n\\nReissing scale\\n\\nReissing scale – not grouped\\n1.43 (1.003–2.046)\\n0.048\\n\\nReissing scale when grouped\\n\\nnormal tone (0)\\nreferencereference\\ndecreased tone \\n\\n(range − 3 to -1)\\n0.53 (0.227–1.220)0.134\\nincreased tone \\n\\n(range 1 to 3)\\n0.37 (0.594–4.078)0.368\\n\\nMultivariable analysis of factors associated with pelvic girdle pain (PGP) shortly postpartum (phase 1)\\n\\ndecreased tone \\n\\n\\n(range − 3 to -1)\\n\\n\\nincreased tone \\n\\n\\n(range 1 to 3)\\n\",\n \"Our study showed that nearly 10% of women were diagnosed with PGP during the first days postpartum, and almost 16% reported similar symptoms 6 weeks later. The mean pain intensity and functional limitations within the first days postpartum were moderate, with values corresponding to mild/low 6 weeks postpartum. The likelihood of experiencing PGP shortly after delivery increased with age and reporting PGP during pregnancy. The doming of the abdominal midline was significantly associated with PGP shortly postpartum only in the univariable analysis. The remaining variables related to diastasis recti or pelvic floor function were not associated with PGP shortly after delivery (24–72 h postpartum).\\nThe reported prevalence of postpartum PGP around 12 weeks after delivery ranges from 3.4 to 43.0% [39–46] with the majority of studies [41–45] showing a higher prevalence than demonstrated in this research. This variation may be due to several reasons. Firstly, diverse diagnoses and terminology were used in the mentioned studies, which could lead to discrepancies in prevalence. In our research, postpartum PGP was defined as a pain that persisted postpartum or occurred within the first weeks after delivery [2]. However, in the study of Stomp van den Berg et al. [44] 25% of the 234 women who had PGP at 12 weeks postpartum had no PGP between 0 and 6 weeks after delivery. Secondly, cultural and ethnic factors can play a role in the processes related to pain perception [47]. Although PGP is prevalent worldwide, it is not recognized by health care systems in some countries. Our previous study has shown that PGP during pregnancy was more common in Norwegian than Polish women [6]. In Norway, PGP is one of the most common causes of sick leave among pregnant women [48]. In Poland, PGP is not commonly recognized, and the term ‘pelvic girdle pain’ is not widely used within health care services. Lower social awareness about this condition could lead to lower reporting. The possible role of ethnicity was noticed in another PGP study [49] indicating a more detailed investigation encompassing cultural and ethnic influences associated with PGP is needed.\\nWe could observe similar discrepancies when analyzing the values related to pain intensity and functional limitations, possibly related to the same reasons as those mentioned above. For instance, in the study by Mukkanavar et al. [41] among Indian postpartum women, as many as 84.5% participants with PGP between the 3rd and 18th week after delivery rated their symptoms as greater than 60 mm on the VAS scale. Stomp van de Berg [44] reported median pain intensity 6 weeks postpartum at 4.3 of NRS scale. In the study of Dunn et al. [42] the mean pain intensity values measured on VAS scale were between 22.5 and 55, depending on the location of PGP and co-existing dysfunctions. Our results seem to be in line with those of Sakamoto et al. [50] who also measured functional limitations with PGQ. In the second day postpartum the mean values were oscillating around 47% (95%CI 40–54), while 4 weeks after − 19% (95%CI 12–25).\\nOur results showing higher PGP prevalence 6 weeks postpartum when compared to early postpartum period may seem contradictory to previous reports [37, 51]. However, it has to be noted that the cited studies followed women experiencing PGP already during pregnancy. The occurrence of pregnancy-related PGP may be delayed up to the first weeks postpartum [2]. By following all women (with and without pain), our study could capture those individuals that developed pain after the initial examination, 24-72 h postpartum. Additionally, early postpartum period is associated with more bed rest when compared to 6 weeks postpartum when PGP symptoms could be more noticeable and bothersome.\\nWhen it comes to factors associated with postpartum PGP, our results are in line with previous reports. A recent systematic review by Wiezer et al. [52] confirmed PGP during pregnancy as a risk factor for persistent postpartum pain. These findings suggest that asking women whether they have experienced PGP during pregnancy may help identify women at risk of persistent postpartum pain. The association between postpartum PGP and age has also been previously shown. Gausel et al. [46] reported age 30 and above as the risk factors for persistent postpartum pain. In European countries, primiparous women are becoming older. In Poland, the mean age of women having their first baby in 2019 was 27.4 and, although constantly increasing, is still one of the lowest in Europe [53]. Increasing maternal age may have several consequences. In accordance with previously mentioned studies, our results indicate that pregnant women who deliver past a certain age should receive special physiotherapy care.\\nIn our study, the doming of the abdominal wall in the projection of linea alba was a statistically significant factor only in univariable analysis and there were no associations between the DRA severity (size of IRD) and the presence of PGP shortly postpartum. This is different when compared to our recently published matched-case control studies [54, 55]. However, in mentioned reports participants were matched according to age and parity, mode of delivery and time postpartum. This may suggest that although DRA features and postpartum PGP may co-exist, those associations are not straightforward and there are possibly other factors that may mediate this relationship.\\nOur study did not reveal any associations between pelvic floor function and PGP shortly postpartum, despite previous reports [17, 55]. This may be due to the timing of phase 1 of the study when the measurements were taken - some differences in the pelvic floor function may be too subtle to be detected using screening palpation examination in the early postpartum period. Our other hypothesis is that there are no differences in the pelvic floor muscle function between women with and without postpartum PGP shortly after delivery. They may be more visible with time, while pain persists and the adaptive changes in the pelvic floor occur, which could be supported by our other PGP study [55].\\n[SUBTITLE] Strengths and limitations [SUBSECTION] This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\\nThis was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\\n[SUBTITLE] Implications [SUBSECTION] Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].\\nObtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].\",\n \"This was the first large-scale study conducted in Poland using the recommended guidelines for classifying and investigating postpartum PGP prevalence with the use of screening palpation examination of the pelvic floor and abdominal muscles. To our knowledge, this is also the first study in the central-eastern region in Europe. Considering that this region is inhabited mainly by Caucasian women with a similar physiognomy, our results may estimate postpartum PGP prevalence in this part of Europe.\\nThe main limitation of this work is the high drop-out rate in phase 2 (“follow-up”). For this reason, the prevalence of PGP 6 weeks postpartum may be underestimated. Time constraints, lack of trust, and low awareness of clinical trials are the main barriers to participation in research projects [56]. Additionally, the first weeks after delivery are challenging for many women, and research obligations may not be their priority. High drop-out rates were also reported by another study investigating PGP 6 weeks postpartum via SMS where the response rate of 43% was recorded 6 weeks postpartum [37]. Another limitation could be caused by the assessment of PGP 6 weeks via self-reports. However, this method was used in previous PGP research [37, 42] and a study by Rejano-Campo et al. [57] showed that self-reported PGP was verified by specific clinical tests in nearly all cases.\\nFinally, we cannot exclude potential selection bias. Although we have made an effort to adopt random recruitment for phase 1 (shortly postpartum), we have included mainly highly educated women from only one center located in the capital city of Poland. This should be taken into account while inferring results from our sample to the general population.\",\n \"Obtained results with regard to other recently published matched case-control studies suggest that the relationship between PGP and DRA-related factors is multidimensional and not straightforward as previously suggested. Assessment of DRA-related factors seems not to be a crucial part of the screening for postpartum PGP but may be of greater importance when assessing individuals with postpartum PGP. Future research should further investigate the possible, multidimensional interactions between PGP and the whole abdominal wall complex (not restricted to only IRD as recommended by Delphi Consensus Study for the conservative management of pregnancy-related DRA [58]), and whether the DRA-related dysfunctions “only” co-exist with PGP or play a role in it. In that case, future studies focusing on creating adequate tension through the abdominal wall during PGP rehabilitation may be feasible. It should all be adjusted for psychosocial factors, which weren’t taken into account in our study. However, they are related to central pain mechanisms observed in individuals with persistent postpartum PGP and could be important factors filling the gaps in our current understanding of postpartum PGP [7, 59].\",\n \"The findings presented in this study suggest that every tenth Polish woman may experience PGP during the first days postpartum and every sixth can report similar symptoms 6 weeks later. The pain intensity and functional limitations tend to subside over time: from moderate pain intensity and functional limitation shortly postpartum to mild/low 6 weeks later. Nevertheless, postpartum PGP should not be ignored, especially in the context of the observed continued increase in chronic pain syndromes and their associated consequences. Older age, PGP during pregnancy, and doming of the abdominal midline at the level of linea alba were associated with the experience of PGP within the first days postpartum. Our study showed no association between pelvic floor function and PGP shortly postpartum. However, this may be due to the chosen methodology (assessment shortly postpartum).\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,"results","discussion",null,null,"conclusion"],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n \"discussion\",\n null,\n null,\n \"conclusion\"\n]"},"keywords":{"kind":"list like","value":["Pelvic girdle pain","Postpartum period","Prevalence","Pelvic floor","Pelvic floor disorders","rectus abdominis"],"string":"[\n \"Pelvic girdle pain\",\n \"Postpartum period\",\n \"Prevalence\",\n \"Pelvic floor\",\n \"Pelvic floor disorders\",\n \"rectus abdominis\"\n]"}}},{"rowIdx":376,"cells":{"title":{"kind":"string","value":"Innovative approach to monitor performance of integrated disease surveillance and response after the Ebola outbreak in Sierra Leone: lessons from the field."},"pmid":{"kind":"string","value":"36266711"},"background_abstract":{"kind":"string","value":"Supervision of healthcare workers improves performance if done in a supportive and objective manner. Regular supervision is a support function of Integrated Disease Surveillance and Response (IDSR) strategy and allows systematic monitoring of IDSR implementation. Starting 2015, WHO and other development partners supported the Ministry of Health and Sanitation (MoHS) to revitalize IDSR in Sierra Leone and to monitor progress through supportive supervision assessments. We report on the findings of these assessments."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"This was a cross-sectional study where six longitudinal assessments were conducted in randomly selected health facilities. Health facilities assessed were 71 in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021. An electronic checklist based on selected core functions of IDSR was developed and uploaded onto tablets using the Open Data Kit (ODK) platform. Supervision teams interviewed health care workers, reviewed documents and made observations in health facilities. Supervision books were used to record feedback and corrective actions. Data from the supervisory visits was downloaded from ODK platform, cleaned and analysed. Categorical data was summarized using frequencies and proportions while means and medians were used for continuous variables. Z test was used to test for differences in proportions."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Completeness of IDSR reporting improved from 84.5% in 2016 to 96% in 2021 (11.5% points; 95% CI 3.6, 21.9; P-value 0.003). Timeliness of IDSR reports improved from 80.3 to 92% (11.7% points; 95% CI 2.4, 22.9; P-value 0.01). There was significant improvement in health worker knowledge of IDSR concepts and tools, in availability of IDSR standard case definition posters and reporting tools and in data analysis practices. Availability of vaccines and temperature monitoring tools in health facilities also improved significantly but some indicators dropped such as availability of IDSR technical guidelines and malaria testing kits and drugs."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Supervision using electronic tool contributed to health systems strengthening through longitudinal tracking of core IDSR indicators and other program indicators such as essential malaria commodities and availability and status of routine vaccines. Supervision using electronic tools should be extended to other programs."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Hemorrhagic Fever, Ebola","Public Health Surveillance","Sierra Leone","Cross-Sectional Studies","Disease Outbreaks"],"string":"[\n \"Humans\",\n \"Hemorrhagic Fever, Ebola\",\n \"Public Health Surveillance\",\n \"Sierra Leone\",\n \"Cross-Sectional Studies\",\n \"Disease Outbreaks\"\n]"},"pmcid":{"kind":"string","value":"9584265"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Coordination of IDSR activities in health facilities [SUBSECTION] IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\nIDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n[SUBTITLE] Identification of priority IDSR diseases, conditions and events [SUBSECTION] The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\nThe proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n[SUBTITLE] Awareness and adherence to IDSR reporting requirements [SUBSECTION] Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\nKnowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\n[SUBTITLE] Availability of IDSR reporting tools [SUBSECTION] Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\nAvailability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\n[SUBTITLE] Data analysis and use at health facility level [SUBSECTION] There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\nThere was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\n[SUBTITLE] Outbreak detection and notification [SUBSECTION] The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\nThe proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\n[SUBTITLE] Feedback, monitoring and communication [SUBSECTION] Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\nQuarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\n[SUBTITLE] Availability of routine immunization services and commodities for management of Malaria [SUBSECTION] The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge\nThe proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge"},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"The IDSR system is now well established in Sierra Leone. The support supervision visits that were done using an integrated electronic tool contributed to health systems strengthening through longitudinal tracking of core IDSR indicators and other program indicators such as essential malaria commodities and availability and status of routine vaccines. The feedback that was provided to all levels of healthcare including the national program and development partners supported to address the gaps identified and hence improved performance over the years. MoHS should entrench the supportive supervision approach by the national and district teams using electronic tools to assure sustained monitoring of IDSR and other programs."},"other_sections_titles":{"kind":"list like","value":["Background","The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone","Improving public health surveillance after the Ebola outbreak","Methods","Study setting and design","Selection of health facilities","Data collection","Data analysis","Coordination of IDSR activities in health facilities","Identification of priority IDSR diseases, conditions and events","Awareness and adherence to IDSR reporting requirements","Availability of IDSR reporting tools","Data analysis and use at health facility level","Outbreak detection and notification","Feedback, monitoring and communication","Availability of routine immunization services and commodities for management of Malaria"],"string":"[\n \"Background\",\n \"The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone\",\n \"Improving public health surveillance after the Ebola outbreak\",\n \"Methods\",\n \"Study setting and design\",\n \"Selection of health facilities\",\n \"Data collection\",\n \"Data analysis\",\n \"Coordination of IDSR activities in health facilities\",\n \"Identification of priority IDSR diseases, conditions and events\",\n \"Awareness and adherence to IDSR reporting requirements\",\n \"Availability of IDSR reporting tools\",\n \"Data analysis and use at health facility level\",\n \"Outbreak detection and notification\",\n \"Feedback, monitoring and communication\",\n \"Availability of routine immunization services and commodities for management of Malaria\"\n]"},"other_sections_texts":{"kind":"list like","value":["Supportive supervision is a unique approach to monitoring performance through face to face interactions between a skilled health care worker and a less skilled worker [1]. First promoted in the era of primary health care, supportive supervision sought to improve performance of healthcare workers offering essential services such as immunization in remote areas [2]. However, this approach can be impeded by poor coordination, lack of motivation, geographical barriers, competing activities, high costs and armed conflicts [1, 3].\nInitially, a top-down approach to supervision was used with inspection and correction of junior workers done by a senior officer. There has been a shift to a more collaborative approach (support supervision) with focus on identifying challenges in service delivery and involving the provider in finding solutions. This approach improves motivation and ownership by the health care providers [4]. Records reviews, direct observations, use of objective indicators, collaborative problem solving and targeted on job training are additional collaborative approaches used [1, 5].\nIn 1998, the World Health Organization (WHO), Africa Regional Office adopted the Integrated Disease Surveillance (IDS) strategy, later renamed Integrated Disease Surveillance and Response (IDSR) strategy to strengthen public health surveillance and response in member states [6]. Implementation of the IDSR strategy varies, with the initial momentum exhibited in the first years of implementation declining over time due to several challenges such as non-sustainable financial resources, poor coordination, inadequate training, high turnover of staff, inadequate supervision from the next level, erratic feedback, weak laboratory capacities, unavailability of job aids and poor availability of communication and transport systems particularly at the periphery [7–9]. Considering the growing threat of epidemics especially from new and emerging pathogens, the need for strengthened surveillance systems has never been greater. Ministries of Health in WHO member states are urged to monitor implementation of IDSR as this directly contributes to fulfillment of the International Health Regulations (2005) [10]. National level health staff utilize supervision visits to monitor the implementation of Integrated Disease Surveillance and Response (IDSR) activities and identify key areas that need improvement [11]. While training of sub- national health care workers on IDSR provides them with basic understanding of public health surveillance, supportive supervision reinforces concepts, and identifies and resolves challenges affecting implementation [12].\n[SUBTITLE] The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone [SUBSECTION] Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\nHealth care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\n[SUBTITLE] Improving public health surveillance after the Ebola outbreak [SUBSECTION] Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\nTowards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.","Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].","Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.","[SUBTITLE] Study setting and design [SUBSECTION] This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\nThis cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\n[SUBTITLE] Selection of health facilities [SUBSECTION] The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\nThe total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n[SUBTITLE] Data collection [SUBSECTION] The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\nThe assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\n[SUBTITLE] Data analysis [SUBSECTION] Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\nData was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.","This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.","The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021","The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.","Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.","IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team","The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021","Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).","Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).","There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).","The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).","Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.","The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge"],"string":"[\n \"Supportive supervision is a unique approach to monitoring performance through face to face interactions between a skilled health care worker and a less skilled worker [1]. First promoted in the era of primary health care, supportive supervision sought to improve performance of healthcare workers offering essential services such as immunization in remote areas [2]. However, this approach can be impeded by poor coordination, lack of motivation, geographical barriers, competing activities, high costs and armed conflicts [1, 3].\\nInitially, a top-down approach to supervision was used with inspection and correction of junior workers done by a senior officer. There has been a shift to a more collaborative approach (support supervision) with focus on identifying challenges in service delivery and involving the provider in finding solutions. This approach improves motivation and ownership by the health care providers [4]. Records reviews, direct observations, use of objective indicators, collaborative problem solving and targeted on job training are additional collaborative approaches used [1, 5].\\nIn 1998, the World Health Organization (WHO), Africa Regional Office adopted the Integrated Disease Surveillance (IDS) strategy, later renamed Integrated Disease Surveillance and Response (IDSR) strategy to strengthen public health surveillance and response in member states [6]. Implementation of the IDSR strategy varies, with the initial momentum exhibited in the first years of implementation declining over time due to several challenges such as non-sustainable financial resources, poor coordination, inadequate training, high turnover of staff, inadequate supervision from the next level, erratic feedback, weak laboratory capacities, unavailability of job aids and poor availability of communication and transport systems particularly at the periphery [7–9]. Considering the growing threat of epidemics especially from new and emerging pathogens, the need for strengthened surveillance systems has never been greater. Ministries of Health in WHO member states are urged to monitor implementation of IDSR as this directly contributes to fulfillment of the International Health Regulations (2005) [10]. National level health staff utilize supervision visits to monitor the implementation of Integrated Disease Surveillance and Response (IDSR) activities and identify key areas that need improvement [11]. While training of sub- national health care workers on IDSR provides them with basic understanding of public health surveillance, supportive supervision reinforces concepts, and identifies and resolves challenges affecting implementation [12].\\n[SUBTITLE] The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone [SUBSECTION] Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\\nHealth care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\\n[SUBTITLE] Improving public health surveillance after the Ebola outbreak [SUBSECTION] Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\\nTowards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\",\n \"Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\",\n \"Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\",\n \"[SUBTITLE] Study setting and design [SUBSECTION] This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\\nThis cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\\n[SUBTITLE] Selection of health facilities [SUBSECTION] The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\nThe total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n[SUBTITLE] Data collection [SUBSECTION] The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\\nThe assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\\n[SUBTITLE] Data analysis [SUBSECTION] Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\\nData was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\",\n \"This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\",\n \"The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\",\n \"The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\",\n \"Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\",\n \"IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\\n\\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\\n< 0.0001\\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\\n0.002\\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\\n< 0.0001\\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\\n< 0.0001\\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\\n0.03\\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\\n0.0008\\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\\n< 0.0001\\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\\n< 0.0001\\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\n\\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\\n*Denominator is number of outbreaks reported in the preceding 12 months\\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\",\n \"The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\\n\\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\n\\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\",\n \"Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\",\n \"Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\",\n \"There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\",\n \"The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\",\n \"Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\",\n \"The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\\n\\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\\nVaccination\\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\\n0.0157\\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\\n< 0.0001\\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\\n< 0.0001\\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\\n0.002\\n\\nMalaria Commodities\\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\\n0.0002\\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\\n0.001\\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\\n0.0218\\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\\n\\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\\n†Denominator is number of health facilities providing immunization services\\n†† Denominator is number of health facilities with a functional fridge\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone","Improving public health surveillance after the Ebola outbreak","Methods","Study setting and design","Selection of health facilities","Data collection","Data analysis","Results","Coordination of IDSR activities in health facilities","Identification of priority IDSR diseases, conditions and events","Awareness and adherence to IDSR reporting requirements","Availability of IDSR reporting tools","Data analysis and use at health facility level","Outbreak detection and notification","Feedback, monitoring and communication","Availability of routine immunization services and commodities for management of Malaria","Discussion","Conclusion"],"string":"[\n \"Background\",\n \"The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone\",\n \"Improving public health surveillance after the Ebola outbreak\",\n \"Methods\",\n \"Study setting and design\",\n \"Selection of health facilities\",\n \"Data collection\",\n \"Data analysis\",\n \"Results\",\n \"Coordination of IDSR activities in health facilities\",\n \"Identification of priority IDSR diseases, conditions and events\",\n \"Awareness and adherence to IDSR reporting requirements\",\n \"Availability of IDSR reporting tools\",\n \"Data analysis and use at health facility level\",\n \"Outbreak detection and notification\",\n \"Feedback, monitoring and communication\",\n \"Availability of routine immunization services and commodities for management of Malaria\",\n \"Discussion\",\n \"Conclusion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Supportive supervision is a unique approach to monitoring performance through face to face interactions between a skilled health care worker and a less skilled worker [1]. First promoted in the era of primary health care, supportive supervision sought to improve performance of healthcare workers offering essential services such as immunization in remote areas [2]. However, this approach can be impeded by poor coordination, lack of motivation, geographical barriers, competing activities, high costs and armed conflicts [1, 3].\nInitially, a top-down approach to supervision was used with inspection and correction of junior workers done by a senior officer. There has been a shift to a more collaborative approach (support supervision) with focus on identifying challenges in service delivery and involving the provider in finding solutions. This approach improves motivation and ownership by the health care providers [4]. Records reviews, direct observations, use of objective indicators, collaborative problem solving and targeted on job training are additional collaborative approaches used [1, 5].\nIn 1998, the World Health Organization (WHO), Africa Regional Office adopted the Integrated Disease Surveillance (IDS) strategy, later renamed Integrated Disease Surveillance and Response (IDSR) strategy to strengthen public health surveillance and response in member states [6]. Implementation of the IDSR strategy varies, with the initial momentum exhibited in the first years of implementation declining over time due to several challenges such as non-sustainable financial resources, poor coordination, inadequate training, high turnover of staff, inadequate supervision from the next level, erratic feedback, weak laboratory capacities, unavailability of job aids and poor availability of communication and transport systems particularly at the periphery [7–9]. Considering the growing threat of epidemics especially from new and emerging pathogens, the need for strengthened surveillance systems has never been greater. Ministries of Health in WHO member states are urged to monitor implementation of IDSR as this directly contributes to fulfillment of the International Health Regulations (2005) [10]. National level health staff utilize supervision visits to monitor the implementation of Integrated Disease Surveillance and Response (IDSR) activities and identify key areas that need improvement [11]. While training of sub- national health care workers on IDSR provides them with basic understanding of public health surveillance, supportive supervision reinforces concepts, and identifies and resolves challenges affecting implementation [12].\n[SUBTITLE] The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone [SUBSECTION] Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\nHealth care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\n[SUBTITLE] Improving public health surveillance after the Ebola outbreak [SUBSECTION] Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\nTowards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.","Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].","Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\n\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\n\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.","[SUBTITLE] Study setting and design [SUBSECTION] This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\nThis cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\n[SUBTITLE] Selection of health facilities [SUBSECTION] The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\nThe total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n[SUBTITLE] Data collection [SUBSECTION] The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\nThe assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\n[SUBTITLE] Data analysis [SUBSECTION] Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\nData was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.","This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.","The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\n\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\n\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021","The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\n\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\n\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.","Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.","[SUBTITLE] Coordination of IDSR activities in health facilities [SUBSECTION] IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\nIDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n[SUBTITLE] Identification of priority IDSR diseases, conditions and events [SUBSECTION] The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\nThe proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n[SUBTITLE] Awareness and adherence to IDSR reporting requirements [SUBSECTION] Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\nKnowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\n[SUBTITLE] Availability of IDSR reporting tools [SUBSECTION] Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\nAvailability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\n[SUBTITLE] Data analysis and use at health facility level [SUBSECTION] There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\nThere was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\n[SUBTITLE] Outbreak detection and notification [SUBSECTION] The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\nThe proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\n[SUBTITLE] Feedback, monitoring and communication [SUBSECTION] Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\nQuarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\n[SUBTITLE] Availability of routine immunization services and commodities for management of Malaria [SUBSECTION] The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge\nThe proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge","IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\n\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\n< 0.0001\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\n0.002\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\n< 0.0001\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\n< 0.0001\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\n0.03\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\n0.0008\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\n< 0.0001\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\n< 0.0001\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\n\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\n*Denominator is number of outbreaks reported in the preceding 12 months\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team","The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\n\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\n\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021","Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).","Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).","There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).","The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).","Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.","The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\n\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\nVaccination\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\n0.0157\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\n< 0.0001\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\n< 0.0001\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\n0.002\n\nMalaria Commodities\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\n0.0002\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\n0.001\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\n0.0218\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\n\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\n†Denominator is number of health facilities providing immunization services\n†† Denominator is number of health facilities with a functional fridge","Through the support supervision data collected and analysed after each visit to the districts, the country was able to monitor trends of key IDSR indicators. In general, health facilities performed well over the years in several assessment areas (Tables 2 and 3). However, there were also some other areas that did not improve much due to several challenges including inadequate funding and COVID-19 pandemic. The funding challenge was partly resolved by sharing the supervision feedback with stakeholders who could provide financial support to address some of the challenges. For example, WHO and other development partners supported the MoHS to print and distribute surveillance tools to health facilities and to conduct refresher trainings for health workers focusing on case definitions of priority diseases, conditions and events listed in the IDSR technical guidelines. However, as the findings of the availability of surveillance tools shows (Table 2), only the weekly reporting tool had become more available over the years compared to the other tools (line listing forms, case-based reporting forms and rumour logbooks). This is because the weekly tool is the one more frequently used compared to the others and hence development partners were more likely to support in its printing.\nSurveillance data is more accurate in measuring burden of disease if it is representative and timely [9]. IDSR report completeness (proportion of health facilities submitting reports) and timeliness improved during the review period and may be partly attributed to availability of IDSR focal persons in most health facilities who were familiar with IDSR reporting obligations and were equipped with the tools necessary for reporting. Lack of IDSR focal persons and unavailability of technical guidelines are often associated with low weekly IDSR reporting rates and timeliness [17]. Gradual migration from paper based to electronic reporting of IDSR data in Sierra Leone at the health facility level starting mid-2017 may have improved timeliness of IDSR reports by reducing report transmission time [18]. Reporting completeness and timeliness rates in these assessments are comparable to those found in a similar assessment in Uganda [19].\nOver the years, there was overall improvement in knowledge of case definitions for priority conditions among health workers which could have been due to increased availability of case definition posters in health facilities. However, the proportion of health workers with adequate knowledge of the case definitions was less than 80% in all the diseases sampled which means that not all health workers were using the case definition posters (Fig. 2). For Ebola Virus Disease, the suboptimal knowledge of the case definition may have been due to changes in case definitions (from outbreak case definition to a routine surveillance case definition). Low knowledge on the case definition for Cholera was likely because the country had not recorded any case since the last outbreak in 2013.\nSyndromic surveillance is important in low resource settings where laboratory confirmation is not always readily available as is currently the case in Sierra Leone. Inadequate knowledge on case definitions leads to low levels of suspicion or misclassification of cases. Hence, it is important that healthcare workers use clinical signs and symptoms as depicted in the standard case definition posters to suspect cases and initiate investigation. The Ministry of Health and Sanitation must therefore do more to engage the health workers in all health facilities to ensure that they read and master the case definitions for the various diseases since the posters are available in most health facilities.\nData analysis at the health facilities improved markedly over time from 2016 to 2021 (Table 2) and the reasons given by the health workers for this improvement were mainly training on data analysis skills, provision of data analysis graphs and tablets for reporting. In deed these are important elements in data analysis as lack of computers and technical capacity has been reported in other countries as being responsible for poor data analysis [19]. Limited laboratory diagnostic capacity was observed even for organisms such as Vibrio cholerae that have caused large magnitude epidemics in Sierra Leone in the past [20]. This limits the contribution of laboratories to outbreak detection, guiding case management and monitoring trends of priority diseases as specimen referral increases turnaround time for results.\nWhile routine vaccination services were available in most health facilities, gaps were observed in maintenance of cold chain that could have compromised the quality of vaccines. It is worth noting that the gaps reduced in subsequent assessments with significant improvement in the number of health facilities with all basic vaccines and functional cold chain equipment. Availability of first line antimalarial drugs and rapid diagnostic kits for malaria detection dropped during the end line assessment in 2021 and this could have been attributed to the COVID-19 pandemic which has caused constraint on available resources including for malaria [21].\nThis study had a few limitations. Even though two national IDSR support supervision visits were planned each year, only one was conducted annually except in 2016 when two were made. This was mostly due to competing priorities or resource constraints. While two visits per year would have provided us with better trends, it did not affect the quality of the data. Support supervision was however conducted by the districts on a quarterly basis although the comprehensive electronic checklist was not always used and this data could therefore not be included. Additionally, due to the COVID-19 pandemic, no national IDSR support supervision was conducted in 2020 due to travel restrictions to the districts for most part of the year. However, there were other visits made to the districts as part of the response to the pandemic.","The IDSR system is now well established in Sierra Leone. The support supervision visits that were done using an integrated electronic tool contributed to health systems strengthening through longitudinal tracking of core IDSR indicators and other program indicators such as essential malaria commodities and availability and status of routine vaccines. The feedback that was provided to all levels of healthcare including the national program and development partners supported to address the gaps identified and hence improved performance over the years. MoHS should entrench the supportive supervision approach by the national and district teams using electronic tools to assure sustained monitoring of IDSR and other programs."],"string":"[\n \"Supportive supervision is a unique approach to monitoring performance through face to face interactions between a skilled health care worker and a less skilled worker [1]. First promoted in the era of primary health care, supportive supervision sought to improve performance of healthcare workers offering essential services such as immunization in remote areas [2]. However, this approach can be impeded by poor coordination, lack of motivation, geographical barriers, competing activities, high costs and armed conflicts [1, 3].\\nInitially, a top-down approach to supervision was used with inspection and correction of junior workers done by a senior officer. There has been a shift to a more collaborative approach (support supervision) with focus on identifying challenges in service delivery and involving the provider in finding solutions. This approach improves motivation and ownership by the health care providers [4]. Records reviews, direct observations, use of objective indicators, collaborative problem solving and targeted on job training are additional collaborative approaches used [1, 5].\\nIn 1998, the World Health Organization (WHO), Africa Regional Office adopted the Integrated Disease Surveillance (IDS) strategy, later renamed Integrated Disease Surveillance and Response (IDSR) strategy to strengthen public health surveillance and response in member states [6]. Implementation of the IDSR strategy varies, with the initial momentum exhibited in the first years of implementation declining over time due to several challenges such as non-sustainable financial resources, poor coordination, inadequate training, high turnover of staff, inadequate supervision from the next level, erratic feedback, weak laboratory capacities, unavailability of job aids and poor availability of communication and transport systems particularly at the periphery [7–9]. Considering the growing threat of epidemics especially from new and emerging pathogens, the need for strengthened surveillance systems has never been greater. Ministries of Health in WHO member states are urged to monitor implementation of IDSR as this directly contributes to fulfillment of the International Health Regulations (2005) [10]. National level health staff utilize supervision visits to monitor the implementation of Integrated Disease Surveillance and Response (IDSR) activities and identify key areas that need improvement [11]. While training of sub- national health care workers on IDSR provides them with basic understanding of public health surveillance, supportive supervision reinforces concepts, and identifies and resolves challenges affecting implementation [12].\\n[SUBTITLE] The effect on Ebola Virus Disease outbreak on the health workforce in Sierra Leone [SUBSECTION] Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\\nHealth care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\\n[SUBTITLE] Improving public health surveillance after the Ebola outbreak [SUBSECTION] Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\\nTowards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\",\n \"Health care workers are a high-risk group for highly infectious diseases such as Ebola Virus Disease (EVD). Out of 304 health care workers infected with Ebola virus in Sierra Leone in 2014–2016, two hundred and twenty-one (221) died and several others resigned out of fear of getting infected. This reduced the ratio of skilled health workers from 17.2/ 10,000 to < 4 /10,000 population [13–15]. After the outbreak, the Ministry of Health and Sanitation (MoHS) embarked on strengthening health systems, improving public health surveillance and cross border surveillance [14].\",\n \"Towards the end of the Ebola Virus Disease outbreak, the Sierra Leone Ministry of Health and Sanitation in collaboration with World Health Organization country office (WCO) and other health sector development partners, embarked on revitalization of IDSR. This involved adaptation of the 2010 WHO-Africa region IDSR technical guidelines and training modules, training of healthcare workers and providing inputs and infrastructure to support IDSR implementation [16].\\n\\nPreviously, paper-based checklists were used during supportive supervision in Sierra Leone and health facilities were selected purposively based on ease of access. Thus, it was likely that health facilities in remote areas were omitted from supervisory visits. Collation and analysis of data from the supervisory visits was seldom done as it required abstraction of data from paper-based checklists to a computer and this often required hiring of data clerks due to staff shortage. To overcome these issues, we adopted an electronic supervision checklist and an open source platform to collect and manage data. This paper describes the innovative, integrated approach to IDSR supportive supervision. We share the lessons learnt, hopeful that they can provide insight for improving monitoring of public health surveillance programs particularly for countries implementing IDSR.\\n\\nThe integrated support supervision process described in this paper is considered an innovative way of monitoring IDSR performance indicators for several reasons. First, routine supervision visits were used to collect comprehensive and representative data to monitor key IDSR indicators from 2016 to 2021. During the visits, verbal and written feedback on performance was then given to the health facilities and district health management teams which ensured that corrective action was taken where necessary. Second, an integrated questionnaire was used which not only collected surveillance data, but also data from other related programs such as laboratory, immunization and malaria. Third, the use of electronic supervision checklist and an open source data collection platform to collect and manage the data ensured secure storage and rapid retrieval of data for analysis which enabled monitoring of performance over time.\",\n \"[SUBTITLE] Study setting and design [SUBSECTION] This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\\nThis cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\\n[SUBTITLE] Selection of health facilities [SUBSECTION] The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\nThe total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n[SUBTITLE] Data collection [SUBSECTION] The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\\nThe assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\\n[SUBTITLE] Data analysis [SUBSECTION] Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\\nData was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\",\n \"This cross-sectional study was part of operational research to monitor implementation of IDSR program in Sierra Leone post Ebola outbreak. Data was collected through supportive supervision visits to health facilities across all districts in Sierra Leone from 2016 to 2021. The data collected was used to monitor and improve IDSR implementation in the country.\",\n \"The total number of health facilities sampled nationally (sample size) per each support supervision visit was purposively selected based on available resources. There were about 1425 health facilities in the country in 2021, distributed by type as follows, in order of size and scope of services offered (from largest to smallest): Hospital 4%, Community Health Centres (CHC) 19%, Community Health Posts (CHP) 30%, Maternal and Child Health Posts (MCHP) 43% and Clinic 3%. About 10% of the total existing facilities per year was sampled in each visit. To ensure representativeness, health facilities were selected from each type of health facility (except clinics which are mostly private) in all the districts. Selection of health facilities in each district was done randomly from a list of health facilities stratified by health facility type as indicated above. First, health facilities were stratified according to the type of health facility. Thereafter, random sampling, using computer generated numbers was done to identify at least one hospital, three CHC, three CHP and three MCHP in each district. Where available, major private health facilities equivalent to a hospital were included.\\nOver the six-year period, six national level supportive supervision visits were conducted covering 71 health facilities in February 2016, 99 in July 2016, 101 in May 2017, 126 in August 2018, 139 in February 2019 and 156 in August 2021 (Fig. 1). No national level supervisory visit was carried out in 2020 due to COVID-19 pandemic. Apart from the national supervision visits, the District Health Management Teams conducted two to four supervision visits per year for the facilities under their jurisdiction. However, this data was captured in a separate database and is therefore not included in this article.\\n\\nFig. 1Map of Sierra Leone showing health facilities visited during support supervision, 2016–2021\\n\\nMap of Sierra Leone showing health facilities visited during support supervision, 2016–2021\",\n \"The assessment teams comprised of national level staff from the surveillance and laboratory programs of MOHS as well as development partners like World Health Organization who were supporting the supervision missions technically and financially. Respondents in the assessments were a team of healthcare workers involved in surveillance and administrative functions in the health facilities. These included the IDSR focal person, health facility in charge, laboratory technician and the nurse in charge of vaccination activities.\\nOne week prior to each supportive supervisory visit, the MoHS through the National Disease Surveillance Program contacted the District Medical Officers (DMO) and District Surveillance Officers (DSOs) to inform them of the upcoming supervisory visits. On the first day of the visit, the supervisory team comprising of officials from MoHS and supporting development partners met with the District Health Management Team (DHMT) members to discuss the purpose of the supervisory visits and to select health facilities for the assessment. The selected health facilities were not informed of the visit in advance in order to avoid pre-emptive interference with data tools. Upon reaching a health facility, the supervisory team captured the geographic coordinates of the health facilities using automatically generated geo- coordinates. Recording of geo-coordinates for each facility ensured that there was evidence of representativeness in the distribution of selected health facilities. Next the team conducted interviews, verified documents and confirmed availability of IDSR reporting tools.\\nData was collected in real-time using a structured electronic questionnaire created using Open Data Kit platform and loaded onto tablets. Using the electronic tool eased data collection and management, making it possible to analyse data, provide timely feedback to the health facilities staff, and also track performance over time. The questionnaire covered six out of eight core functions of IDSR as well as an assessment on related programmatic areas such as available laboratory services, maternal deaths, malaria commodities and routine immunization services. The questionnaire collected information on (1) case identification; (2) reporting of priority diseases; (3) data analysis and interpretation; (4) investigation and confirmation of reported cases and outbreaks; (5) feedback between the district management teams and health facilities; and (6) monitoring and evaluation (Table 1).\\n\\nTable 1IDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021Core FunctionAssessment methodsMain focus areas1Identification of priority diseases, conditions and eventsInterview, observation, review of registersAssessed the ability of healthcare workers to use standard case definition to identify priority diseases through records review and active case search in the community2Reporting of Priority diseasesInterview, documents review, observationChecked if HCWs are aware of reporting requirements for priority diseases and the availability and use of reporting tools3Data analysis and interpretationInterview and observationChecked if data analysis and interpretation was done at the facility4Investigation and confirmation of outbreaksInterviewAssessed if HF are able to detect outbreaks from various sources, notify the district and participate in outbreak investigations5FeedbackInterviewConfirmed if the HF gets feedback on IDSR performance from the DHMT6Monitoring and EvaluationInterview, document reviewAssessed the frequency of supervision done by the DHMT and mechanisms for monitoring IDSR performance in the HFIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\n\\nIDSR Core Functions Assessed During Integrated Support Supervision, Sierra Leone, 2016–2021\\nIDSR: Integrated Disease Surveillance and Response; HCW: Healthcare workers; HF: Health Facility; DHMT: District Health Management Team\\nThe teams began by assessing availability of IDSR focal persons in the health facilities and their participation in district monthly review meetings. This is because the focal persons are important in coordinating IDSR activities. On case identification, supervisors checked if IDSR posters with case definitions were available and displayed in the consultation rooms. They then interviewed the respondents to gauge their knowledge of case definitions for five selected IDSR priority diseases/conditions (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in the 2021 assessment. A healthcare worker who mentioned all signs and symptoms correctly was classified as having adequate knowledge, while one who mentioned some of the signs and symptoms was classified as having limited knowledge. A healthcare worker who did not mention any of the appropriate signs and symptoms was classified as having no knowledge.\\nThe section on reporting of priority diseases assessed knowledge of and compliance with reporting requirements for priority diseases. Copies of IDSR weekly reports for the previous four weeks were checked to confirm if all reports were available and if they had been submitted on time. The team also checked the availability of case-based reporting forms, weekly reporting forms, line listing forms, and rumour logbooks for recording suspected outbreaks/events. The section on data analysis assessed the capacity and type of data analysis conducted to monitor trends of priority diseases at the health facilities. The section on investigation captured information on case investigation and confirmation of reported outbreaks. The section on feedback assessed feedback mechanisms from the district to the health facilities and vice versa. It included a review of previous supervision reports to elicit frequency of supportive supervision by DHMT. The section on monitoring and evaluation assessed mechanisms for monitoring IDSR performance in the health facility.\\nAdditional questions on immunization, malaria treatment commodities and maternal mortality were also included in the questionnaire. The questionnaire had compulsory fields and intrinsic data validation mechanisms to enhance data accuracy and completeness. Finally, at the end of the data collection process, the teams used supervision books available in each health facility to record strengths, gaps, challenges and recommendations agreed upon during the supervisory visit. These records were useful for follow up during subsequent supervisory visits. Supervision also provided an opportunity for on the job training and distribution of IDSR reference materials.\",\n \"Data was extracted from the Open Data Kit (ODK) platform first onto an MS Excel database, merged, checked for duplicates and completeness and then exported to Epi Info for analysis. We calculated means and medians for continuous variables and proportions for categorical data. Two proportion Z-test was used to compare differences in performance recorded during the baseline in February 2016 and the end line in August 2021. District weekly report completeness was defined as the proportion of health facilities that submitted required reports to the district health office. This was verified by checking if all copies of the IDSR weekly reports for the preceding four weeks were available. Timeliness was defined as the proportion of health facilities that submitted IDSR weekly reports to the district by 12 pm every Monday.\",\n \"[SUBTITLE] Coordination of IDSR activities in health facilities [SUBSECTION] IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\\n\\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\\n< 0.0001\\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\\n0.002\\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\\n< 0.0001\\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\\n< 0.0001\\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\\n0.03\\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\\n0.0008\\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\\n< 0.0001\\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\\n< 0.0001\\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\n\\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\\n*Denominator is number of outbreaks reported in the preceding 12 months\\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\nIDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\\n\\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\\n< 0.0001\\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\\n0.002\\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\\n< 0.0001\\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\\n< 0.0001\\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\\n0.03\\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\\n0.0008\\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\\n< 0.0001\\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\\n< 0.0001\\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\n\\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\\n*Denominator is number of outbreaks reported in the preceding 12 months\\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\n[SUBTITLE] Identification of priority IDSR diseases, conditions and events [SUBSECTION] The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\\n\\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\n\\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\nThe proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\\n\\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\n\\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\n[SUBTITLE] Awareness and adherence to IDSR reporting requirements [SUBSECTION] Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\\nKnowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\\n[SUBTITLE] Availability of IDSR reporting tools [SUBSECTION] Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\\nAvailability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\\n[SUBTITLE] Data analysis and use at health facility level [SUBSECTION] There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\\nThere was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\\n[SUBTITLE] Outbreak detection and notification [SUBSECTION] The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\\nThe proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\\n[SUBTITLE] Feedback, monitoring and communication [SUBSECTION] Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\\nQuarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\\n[SUBTITLE] Availability of routine immunization services and commodities for management of Malaria [SUBSECTION] The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\\n\\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\\nVaccination\\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\\n0.0157\\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\\n< 0.0001\\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\\n< 0.0001\\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\\n0.002\\n\\nMalaria Commodities\\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\\n0.0002\\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\\n0.001\\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\\n0.0218\\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\\n\\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\\n†Denominator is number of health facilities providing immunization services\\n†† Denominator is number of health facilities with a functional fridge\\nThe proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\\n\\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\\nVaccination\\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\\n0.0157\\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\\n< 0.0001\\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\\n< 0.0001\\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\\n0.002\\n\\nMalaria Commodities\\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\\n0.0002\\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\\n0.001\\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\\n0.0218\\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\\n\\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\\n†Denominator is number of health facilities providing immunization services\\n†† Denominator is number of health facilities with a functional fridge\",\n \"IDSR focal persons were present in most health facilities during all assessments. When compared to the baseline assessment, the proportion of IDSR focal persons who participated in the monthly district management team meetings increased significantly from 74.6% to 2016 (baseline) to 96.1% at end line in 2021 (increase of 21.5% points; 95% CI (11.8, 32.9); P-value < 0.0001). Conversely, availability of IDSR technical guidelines at the health facilities declined from 97.2% to 2016 to 82.1% in 2021 (decrease of 15.1% points; 95% CI (6.5, 22.2); P-value 0.002) (Table 2). The decline in availability of IDSR technical guidelines was mainly due to loss without replacement.\\n\\nTable 2Comparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021Core FunctionIndicatorsFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP valueN = 71n (%)N = 99n (%)N = 101n (%)N = 126n (%)N = 139n (%)N = 156n (%)Difference (95 CI)CoordinationHealth facility (HF) has IDSR focal person69 (97.2)97 (98.0)100 (99.0)121 (96.0)135 (97.1)153 (98.1)0.9 (-3.2, 7.9)0.67HF IDSR focal person attends district monthly meetings53 (74.6)87 (87.9)93 (92.1)118 (93.7)132 (95.0)150 (96.1)21.5 (11.8, 32.9)\\n< 0.0001\\nHF has IDSR Technical guidelines69 (97.2)89 (89.9)101 (100.0)102 (81.0)123 (88.5)128 (82.1)15.1 (6.5, 22.2)\\n0.002\\nCase IdentificationStandard Case definition poster displayed in at least one location in the health facility54 (76.1)84 (84.8)89 (88.1)114 (90.5)127 (91.4)148 (94.9)18.8 (9.2, 30.2)\\n< 0.0001\\nIDSR focal person conducts active case search at least once a week56 (78.9)54 (54.5)34 (33.7)100 (79.4)75 (54.0)123 (78.8)0.1 (-12.1, 10.7)0.99Reporting of Priority diseasesHCW correctly defines epidemiologic week51 (71.8)93 (93.9)99 (98.0)108 (85.7)132 (95)151 (96.8)25 (15.0, 36.5)\\n< 0.0001\\nHCW correctly defines zero reporting67 (94.4)94 (94.9)99 (98.0)121 (96.0)138 (99.3)147 (92.2)2.2 (-6.4, 8.5))0.55HCW knows the deadline for submitting weekly IDSR reports65 (91.5)96 (97.0)100 (99.0)115 (91.3)138 (99.3)151 (96.8)5.3 (-0.8, 14.3)0.09Weekly reporting forms available in HF63 (88.7)89 (89.9)96 (95.0)119 (94.4)127 (91.4)150 (96.1)7.4 (0.4, 17.1)\\n0.03\\nLine listing forms available in HF60 (84.5)78 (78.8)81 (80.2)95 (75.4)114 (82.0)122 (78.2)6.3 (-5.4, 16.0)0.27Case based forms available in HF66 (93.0)79 (79.8)91 (90.1)110 (87.3)114 (82.0)139 (89.1)3.9 (-5.4, 11.0)0.36Rumor logs available in HF39 (54.9)55 (55.6)52 (51.5)78 (61.9)84 (60.4)97 (62.1)7.2 (-6.3, 20.8)0.31Data Analysis and UseHF conducts data basic data analysis28 (39.4)49 (49.5)41 (40.6)51 (40.5)63 (45.3)99 (63.4)24.0 (10.0, 36.7)\\n0.0008\\nHF have current line graphs showing trends of priority diseases5 (7.0)11 (11.1)16 (15.8)29 (23.0)43 (30.9)72(46.1)39.1 (27.8, 47.9)\\n< 0.0001\\nOutbreak notification and investigationHF reported outbreak within 12 months12 (16.9)45 (45.5)32 (31.7)10 (7.9)7 (5.0)17(10.9)6.0 (-4.2, 16.1)0.31Outbreak notified to DHMT within 48 h*11 (91.7)29 (64.4)29 (90.6)9 (90.0)5 (71.4)15(88.2%)3.5 (-24.7, 27.1)0.09Monitoring and CommunicationSupervisory visits from DHMT (at least once every three months)44 (62.0)91 (91.9)78 (77.2)79 (62.7)70 (50.4)144 (92.3)30.3(18.6, 42.4)\\n< 0.0001\\nMobile network available in HF67 (94.4)79 (79.8)91 (90.1)120 (95.2)126 (90.6)144 (92.3)2.1 (-6.5, 8.4)0.57*Denominator is number of outbreaks reported in the preceding 12 monthsHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\\n\\nComparison of IDSR indicators in selected health facilities, Sierra Leone, 2016–2021\\n*Denominator is number of outbreaks reported in the preceding 12 months\\nHF: Health Facility; IDSR: Integrated Disease Surveillance and Response; HCW: Healthcare worker; DHMT: District Health Management Team\",\n \"The proportion of health facilities that displayed standard case definition posters in consultation rooms for use by health workers in identifying cases increased significantly from 76.1% to 2016 to 94.9% in 2021 (increase of 18.8% points; 95% CI 9.2, 30.2); P- value 0.002). The proportion of IDSR focal persons who conducted weekly active case search for priority diseases in the health facilities remained the same at 79% for both baseline and end line although it fluctuated across the years. There was positive trend in adequate knowledge of standard case definitions among health workers for five priority diseases that were assessed as shown in Fig. 2 (Acute flaccid paralysis, Neonatal tetanus, Measles, Cholera and Ebola Virus Disease). Additionally, COVID-19 was added in 2021 and 61% of the health workers had adequate knowledge on its case definition.\\n\\nFig. 2Knowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\\n\\nKnowledge of standard case definitions among interviewed healthcare workers, Sierra Leone, 2016 and 2021\",\n \"Knowledge of IDSR reporting requirements was high and by the end line assessment in 2021, almost all interviewed health care workers correctly defined the epidemiologic week, zero reporting and reporting deadlines (Table 2). There was a significant improvement in the proportion of health facilities that submitted all the required surveillance reports (completeness of reporting) from 84.5% to 2016 to 96% in 2021(increase of 11.5% points; 95% CI 3.6, 21.9; P-value 0.003). During the same period, timeliness of IDSR reports improved from 80.3 to 92% (increase of 11.7% points; 95% CI 2.4, 22.9; P-value 0.01).\",\n \"Availability of the weekly IDSR reporting tool improved from 88.7% to 2016 to 96.1% in 2021 (increase of 7.4% points; 95% CI 0.4, 17.1; P-value 0.03). Availability of other reporting tools did not change significantly and were at 78%, 89% and 62% in 2021 for line listing forms, case-based reporting forms and rumour logbooks, respectively (Table 2).\",\n \"There was significant improvement in data analysis and use in health facilities over the years. The proportion of health facilities that conducted basic data analysis improved from 39% at baseline to 63% at end line (increase of 24% points; 95% CI 10.0, 36.7; P-value 0.0008). The proportion of health facilities with current line graphs showing trends in occurrence of priority diseases increased significantly from 7 to 46.1% (increase of 39.1% points; 95% CI 27.8, 47.9; P-value < 0.0001) (Table 2).\",\n \"The proportion of health facilities that had identified an outbreak within 12 months of the assessment did not change significantly and was 16.9% at baseline compared to 10.9% at end line (P-value 0.31). This was mostly because there were no outbreaks to detect and not for lack of detection capacity. The proportion of identified outbreaks that were notified to the District Health Management Teams within 48 h did not change significantly and were 91.7% at baseline compared to 88.2% at end line (P-value 0.09) (Table 2).\",\n \"Quarterly supervisory visits by DHMTs to the health facilities improved from 62% at baseline to 92% (P-value < 000.1) at end line although it fluctuated for other visits (Table 2). Mobile network connectivity remained high throughout the period and 92% of the health facilities assessed at end line had connectivity.\",\n \"The proportion of health facilities providing routine immunization services was above 90% in all the assessments with the end line being 96%. The number of immunizing health facilities with functional refrigerators significantly improved from 49.2% at baseline to 66.7% (P-value 0.02) although it fluctuated for other visits (Table 3). Immunizing health facilities with updated temperature monitoring charts significantly improved from 34 to 81% (P-value < 0.0001) while health facilities with all required basic antigens (vaccines) available at the time of the assessment improved from 48 to 83% (P-value < 0.0001). At the end line assessment, there were only 6.7% health facilities with Vaccine Vial Monitor (VVM) at stage three or four which is considered as excessive exposure to high temperatures (Table 3). Malaria management commodities including first line anti-malarial drugs, rapid diagnostic test kits and insecticide treated nets were available in most of the health facilities throughout the assessment period although there was a significant drop for first line anti-malarial drugs and rapid diagnostic test kits at end line assessment in August 2021, mostly due to disruptions associated with COVID-19 (Table 3).\\n\\nTable 3Assessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021Programmatic AreaVariableFeb_2016July_2016May_2017Aug_2018Feb_2019Aug_2021End Line vs. BaselineP-value(N = 71)n (%)(N = 99)n (%)(N = 101)n (%)(N = 126)n (%)(N = 139)n (%)(N = 156)n (%)Difference (95 CI)\\nVaccination\\nHealth facilities providing routine Immunization Services65 (91.5)95 (96.0)96 (95.0)120 (95.2)135 (97.1)150 (96.2)4.7 (-1.5, 13.7)0.14Health facilities with functional refrigerator†32 (49.2)70 (73.7)72 (75.0)71 (59.2)88 (65.2)100 (66.7)17.5 (3.3, 31.2)\\n0.0157\\nHealth facilities with updated temperature monitoring chart††11 (34.4)16 (22.9)32 (44.4)29 (40.8)37 (42.0)81 (81)46.6 (27.2,62.0)\\n< 0.0001\\nHealth facilities with all required basic antigens available at the time of the assessment†31(47.7)55(57.9)60(62.5)86 (71.7)103 (76.3)125 (83.3)35.6 (21.9, 48.3)\\n< 0.0001\\nHeath facilities with expired antigens†2 (3.1)3 (3.2)1 (1)1 (0.8)2 (1.5)2 (1.3)1.8 (-2.3, 9.3)0.37Health facilities with Vaccine Vial Monitor at stage three and four†14 (21.5)16 (16.8)14 (14.6)8 (6.7)9 (6.7)10 (6.7)14.8 (5.1, 26.6)\\n0.002\\n\\nMalaria Commodities\\nHealth facilities with 1st line anti-malaria drugs69 (97.2)96 (97.0)99 (98.0)120 (95.2)135 (97.1)120 (76.9)20.3 (11.2, 27.7)\\n0.0002\\nHealth Facilities with rapid diagnostic kits for diagnosis of malaria67 (94.4)89 (89.9)98 (97.0)120 (95.2)134 (96.4)119 (76.3)18.1 (8.1, 26.1)\\n0.001\\nHealth facilities with insecticide treated nets55 (77.5)92 (92.9)97 (96.0)111 (88.1)132 (95.0)139 (89.1)11.6 (1.6, 23.3)\\n0.0218\\n†Denominator is number of health facilities providing immunization services†† Denominator is number of health facilities with a functional fridge\\n\\nAssessing routine immunization, malaria commodities and diagnostic capacity in health facilities, Sierra Leone, 2016–2021\\n†Denominator is number of health facilities providing immunization services\\n†† Denominator is number of health facilities with a functional fridge\",\n \"Through the support supervision data collected and analysed after each visit to the districts, the country was able to monitor trends of key IDSR indicators. In general, health facilities performed well over the years in several assessment areas (Tables 2 and 3). However, there were also some other areas that did not improve much due to several challenges including inadequate funding and COVID-19 pandemic. The funding challenge was partly resolved by sharing the supervision feedback with stakeholders who could provide financial support to address some of the challenges. For example, WHO and other development partners supported the MoHS to print and distribute surveillance tools to health facilities and to conduct refresher trainings for health workers focusing on case definitions of priority diseases, conditions and events listed in the IDSR technical guidelines. However, as the findings of the availability of surveillance tools shows (Table 2), only the weekly reporting tool had become more available over the years compared to the other tools (line listing forms, case-based reporting forms and rumour logbooks). This is because the weekly tool is the one more frequently used compared to the others and hence development partners were more likely to support in its printing.\\nSurveillance data is more accurate in measuring burden of disease if it is representative and timely [9]. IDSR report completeness (proportion of health facilities submitting reports) and timeliness improved during the review period and may be partly attributed to availability of IDSR focal persons in most health facilities who were familiar with IDSR reporting obligations and were equipped with the tools necessary for reporting. Lack of IDSR focal persons and unavailability of technical guidelines are often associated with low weekly IDSR reporting rates and timeliness [17]. Gradual migration from paper based to electronic reporting of IDSR data in Sierra Leone at the health facility level starting mid-2017 may have improved timeliness of IDSR reports by reducing report transmission time [18]. Reporting completeness and timeliness rates in these assessments are comparable to those found in a similar assessment in Uganda [19].\\nOver the years, there was overall improvement in knowledge of case definitions for priority conditions among health workers which could have been due to increased availability of case definition posters in health facilities. However, the proportion of health workers with adequate knowledge of the case definitions was less than 80% in all the diseases sampled which means that not all health workers were using the case definition posters (Fig. 2). For Ebola Virus Disease, the suboptimal knowledge of the case definition may have been due to changes in case definitions (from outbreak case definition to a routine surveillance case definition). Low knowledge on the case definition for Cholera was likely because the country had not recorded any case since the last outbreak in 2013.\\nSyndromic surveillance is important in low resource settings where laboratory confirmation is not always readily available as is currently the case in Sierra Leone. Inadequate knowledge on case definitions leads to low levels of suspicion or misclassification of cases. Hence, it is important that healthcare workers use clinical signs and symptoms as depicted in the standard case definition posters to suspect cases and initiate investigation. The Ministry of Health and Sanitation must therefore do more to engage the health workers in all health facilities to ensure that they read and master the case definitions for the various diseases since the posters are available in most health facilities.\\nData analysis at the health facilities improved markedly over time from 2016 to 2021 (Table 2) and the reasons given by the health workers for this improvement were mainly training on data analysis skills, provision of data analysis graphs and tablets for reporting. In deed these are important elements in data analysis as lack of computers and technical capacity has been reported in other countries as being responsible for poor data analysis [19]. Limited laboratory diagnostic capacity was observed even for organisms such as Vibrio cholerae that have caused large magnitude epidemics in Sierra Leone in the past [20]. This limits the contribution of laboratories to outbreak detection, guiding case management and monitoring trends of priority diseases as specimen referral increases turnaround time for results.\\nWhile routine vaccination services were available in most health facilities, gaps were observed in maintenance of cold chain that could have compromised the quality of vaccines. It is worth noting that the gaps reduced in subsequent assessments with significant improvement in the number of health facilities with all basic vaccines and functional cold chain equipment. Availability of first line antimalarial drugs and rapid diagnostic kits for malaria detection dropped during the end line assessment in 2021 and this could have been attributed to the COVID-19 pandemic which has caused constraint on available resources including for malaria [21].\\nThis study had a few limitations. Even though two national IDSR support supervision visits were planned each year, only one was conducted annually except in 2016 when two were made. This was mostly due to competing priorities or resource constraints. While two visits per year would have provided us with better trends, it did not affect the quality of the data. Support supervision was however conducted by the districts on a quarterly basis although the comprehensive electronic checklist was not always used and this data could therefore not be included. Additionally, due to the COVID-19 pandemic, no national IDSR support supervision was conducted in 2020 due to travel restrictions to the districts for most part of the year. However, there were other visits made to the districts as part of the response to the pandemic.\",\n \"The IDSR system is now well established in Sierra Leone. The support supervision visits that were done using an integrated electronic tool contributed to health systems strengthening through longitudinal tracking of core IDSR indicators and other program indicators such as essential malaria commodities and availability and status of routine vaccines. The feedback that was provided to all levels of healthcare including the national program and development partners supported to address the gaps identified and hence improved performance over the years. MoHS should entrench the supportive supervision approach by the national and district teams using electronic tools to assure sustained monitoring of IDSR and other programs.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,"results",null,null,null,null,null,null,null,null,"discussion","conclusion"],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n \"conclusion\"\n]"},"keywords":{"kind":"list like","value":["Public Health","Integrated Disease Surveillance and Response","Monitoring","Technology"],"string":"[\n \"Public Health\",\n \"Integrated Disease Surveillance and Response\",\n \"Monitoring\",\n \"Technology\"\n]"}}},{"rowIdx":377,"cells":{"title":{"kind":"string","value":"Hyperuricaemia and its association with other risk factors for cardiovascular diseases: A population-based study."},"pmid":{"kind":"string","value":"36266776"},"background_abstract":{"kind":"string","value":"Cardiovascular diseases are very common in the general population, and several factors play a role in their development. The purpose of this study was to investigate the relationship between hyperuricaemia and other cardiovascular disease risk factors."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"This cross-sectional study was conducted on 1008 people over the 15-year-old general population in Kerman, Iran. The blood samples of all patients were analysed for the uric acid serum level, and they completed a checklist including physical activity, previous history of hypertension and diabetes, smoking and opium."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"A number of 1008 cases of people were entered into the study. According to the results of this study, 254 patients had uric acid levels above the 75th percentile (6 mg/dl in males, and 5 mg/dl in females). No significant difference was observed between gender (p = .249) and age groups (p = .125) of people with and without hyperuricaemia. The prevalence of overweight/obesity (p < .001), hypertension (p = .004) and low physical activity (p = .033) was significantly higher in patients with hyperuricaemia. The duration of hypertension was significantly higher in hyperuricaemic individuals (p = .022). Overweight/obesity (OR = 2.67; 95% CI = 1.87-3.82) and hypertension (OR = 1.40; 95% CI = 1.02-1.93) were two significant independent factors that contributed to the increased risk of hyperuricaemia in the subjects."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The uric acid serum level is higher in people with hypertension and overweight/obesity. Hyperuricaemia increases the risk of cardiovascular events, which can be prevented by determining the appropriate strategy for the early diagnosis and treatment of this metabolic disorder."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Male","Female","Humans","Adolescent","Hyperuricemia","Uric Acid","Cardiovascular Diseases","Cross-Sectional Studies","Overweight","Risk Factors","Hypertension","Obesity"],"string":"[\n \"Male\",\n \"Female\",\n \"Humans\",\n \"Adolescent\",\n \"Hyperuricemia\",\n \"Uric Acid\",\n \"Cardiovascular Diseases\",\n \"Cross-Sectional Studies\",\n \"Overweight\",\n \"Risk Factors\",\n \"Hypertension\",\n \"Obesity\"\n]"},"pmcid":{"kind":"string","value":"9659661"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"A number of 1008 cases were entered into the study. The relationship between the studied variables and the prevalence of hyperuricaemia is shown in Table 1. According to the results of this study, 254 patients had uric acid levels above the 75th percentile (more than 6 mg/dl in males, and 5 mg/dl in females). No significant difference was seen between gender (p = .249) and age groups (p = .125) of people with and without hyperuricaemia. The prevalence of overweight/obesity was significantly higher in patients with hyperuricaemia (p < .001). The prevalence of low physical activity in patients with hyperuricaemia was significantly higher than in those with normal uric acid levels (p = .033). The prevalence of smoking (p = 406), tobacco consumption (p = .669) and opium addiction (p = .261) in people with and without hyperuricaemia did not show a significant difference. The frequency of hypertension was significantly higher in people with high uric acid levels compared to people with normal uric acid levels (31.9% and 22.8%, respectively; p = .004). The duration of hypertension was significantly higher in hyperuricaemic individuals (p = .022). The prevalence of diabetes (p = .794) and the duration of diabetes (p = .371) in people with and without hyperuricaemia did not show a significant difference. The duration of hypertension and diabetes was converted into categorized variables according to their median (48 and 72 months, respectively). The prevalence of a longer period of hypertension (more than 48 months) in people with hyperuricaemia was significantly higher than in those without hyperuricaemia (p = .025). The prevalence of a longer duration of diabetes (more than 72 months) between the two groups of people with and without hyperuricaemia was similar (p = .986).\nThe relationship between the studied variables and the presence of hyperuricaemia\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h)).\nChi‐square test.\nMann–Whitney test.\nIn order to determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used on variables whose frequency was significantly different in individuals with and without hyperuricaemia (Table 2). Based on the univariate logistic regression model, the presence of overweight/obesity, low physical activity and hypertension increased the chances of hyperuricaemia in the subjects by 2.85, 1.91, compared to those with intense physical activity and 1.58, respectively (p < .001, p = .019 and p = .004, respectively). According to the results of a multivariate analysis on the three variables, the amount of physical activity did not have a significant effect on increasing the chances of hyperuricaemia and the two variables of overweight/obesity (OR = 2.67; 95% CI = 1.87–3.82) and hypertension (OR = 1.40; 95% CI = 1.02–1.93) significantly increased the chances of hyperuricaemia in the subjects.\nUnivariate and multivariate logistic regression models for identifying determinants of hyperuricaemia\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h))."},"conclusions_title":{"kind":"string","value":"CONCLUSION"},"conclusions_text":{"kind":"string","value":"Our study showed that two independent factors for the development of hyperuricaemia are hypertension and overweight/obesity. Since hyperuricaemia is one of the influential factors in cardiovascular events, these events can be prevented by determining the appropriate strategy for early diagnosis and treatment of this metabolic disorder.\nConsidering that our study was conducted in an urban population and environmental and genetic factors can also be involved in the relationship between hyperuricaemia and other cardiovascular disease risk factors, cohort studies with a larger sample size are needed in this field."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Sample size estimation","Statistical analysis","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","ETHIC STATEMENT"],"string":"[\n \"INTRODUCTION\",\n \"Sample size estimation\",\n \"Statistical analysis\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"ETHIC STATEMENT\"\n]"},"other_sections_texts":{"kind":"list like","value":["Cardiovascular diseases are one of the most important causes of death in the world, and their pathogenesis is such that they are affected by several risk factors such as obesity, diabetes and hypertension.\n1\n Increased serum uric acid levels can lead to endothelial dysfunction resulting in atherosclerosis by releasing reactive oxygen species and is reported as one of the risk factors for cardiovascular diseases.\n2\n\n\nIn the process of purine decomposition, uric acid is the final product.\n3\n A condition in which serum uric acid concentration is high but there are no signs or symptoms of uric acid deposition is called asymptomatic hyperuricaemia. Although gout can occur in a person with any rate of uric acid, about two‐thirds of hyperuricaemic people remain asymptomatic.\n4\n Different studies in different races have had different results on the prevalence of hyperuricaemia in people's age and gender.\n5\n, \n6\n\n\nPrevalence of hyperuricaemia is different in viewpoints of race, age and gender groups, and this serum factor has been mentioned as a risk factor for coronary artery diseases; in this study, the relationship between cardiovascular risk factors and hyperuricaemia was investigated.","To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\n9\n and considering Z = 1.96, p = .21, and D = 0.03:\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\n\n\nTo increase the power of our regression, we investigated 1008 people in our study.","To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.","Farzaneh Yazdi, Mohammad Reza Shakibi, Hamid Najafipour, Amir Baniasad, Mohammad Javad Najafzadeh and Samira Sistani designed the study and collected data, Amir Baniasad, Mohammad Javad Najafzadeh wrote the first draft of the paper and submitted the manuscript, Fatemeh Yazdi contributed to writing and revision of the manuscript. All authors contributed to finalizing the manuscript.","This research project was funded by the Kerman University of Medical Sciences.","The study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). Informed consent was obtained from all participants in the study."],"string":"[\n \"Cardiovascular diseases are one of the most important causes of death in the world, and their pathogenesis is such that they are affected by several risk factors such as obesity, diabetes and hypertension.\\n1\\n Increased serum uric acid levels can lead to endothelial dysfunction resulting in atherosclerosis by releasing reactive oxygen species and is reported as one of the risk factors for cardiovascular diseases.\\n2\\n\\n\\nIn the process of purine decomposition, uric acid is the final product.\\n3\\n A condition in which serum uric acid concentration is high but there are no signs or symptoms of uric acid deposition is called asymptomatic hyperuricaemia. Although gout can occur in a person with any rate of uric acid, about two‐thirds of hyperuricaemic people remain asymptomatic.\\n4\\n Different studies in different races have had different results on the prevalence of hyperuricaemia in people's age and gender.\\n5\\n, \\n6\\n\\n\\nPrevalence of hyperuricaemia is different in viewpoints of race, age and gender groups, and this serum factor has been mentioned as a risk factor for coronary artery diseases; in this study, the relationship between cardiovascular risk factors and hyperuricaemia was investigated.\",\n \"To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\\n9\\n and considering Z = 1.96, p = .21, and D = 0.03:\\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\\n\\n\\nTo increase the power of our regression, we investigated 1008 people in our study.\",\n \"To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.\",\n \"Farzaneh Yazdi, Mohammad Reza Shakibi, Hamid Najafipour, Amir Baniasad, Mohammad Javad Najafzadeh and Samira Sistani designed the study and collected data, Amir Baniasad, Mohammad Javad Najafzadeh wrote the first draft of the paper and submitted the manuscript, Fatemeh Yazdi contributed to writing and revision of the manuscript. All authors contributed to finalizing the manuscript.\",\n \"This research project was funded by the Kerman University of Medical Sciences.\",\n \"The study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). Informed consent was obtained from all participants in the study.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","MATERIALS AND METHODS","Sample size estimation","Statistical analysis","RESULTS","DISCUSSION","CONCLUSION","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","CONFLICT OF INTEREST","ETHIC STATEMENT"],"string":"[\n \"INTRODUCTION\",\n \"MATERIALS AND METHODS\",\n \"Sample size estimation\",\n \"Statistical analysis\",\n \"RESULTS\",\n \"DISCUSSION\",\n \"CONCLUSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"CONFLICT OF INTEREST\",\n \"ETHIC STATEMENT\"\n]"},"all_sections_texts":{"kind":"list like","value":["Cardiovascular diseases are one of the most important causes of death in the world, and their pathogenesis is such that they are affected by several risk factors such as obesity, diabetes and hypertension.\n1\n Increased serum uric acid levels can lead to endothelial dysfunction resulting in atherosclerosis by releasing reactive oxygen species and is reported as one of the risk factors for cardiovascular diseases.\n2\n\n\nIn the process of purine decomposition, uric acid is the final product.\n3\n A condition in which serum uric acid concentration is high but there are no signs or symptoms of uric acid deposition is called asymptomatic hyperuricaemia. Although gout can occur in a person with any rate of uric acid, about two‐thirds of hyperuricaemic people remain asymptomatic.\n4\n Different studies in different races have had different results on the prevalence of hyperuricaemia in people's age and gender.\n5\n, \n6\n\n\nPrevalence of hyperuricaemia is different in viewpoints of race, age and gender groups, and this serum factor has been mentioned as a risk factor for coronary artery diseases; in this study, the relationship between cardiovascular risk factors and hyperuricaemia was investigated.","This study was a sub‐analyse of collected data about the risk factors of coronary artery diseases (i.e., Kerman Coronary Artery Disease Risk Factor Study; KERCADRS). This cross‐sectional and descriptive study was conducted in Kerman city on 1008 people who were over 15 years old. The sampling method of the studied people was in the form of a cluster from the entire population of the city. These people were selected from 720,000 Kerman residents, which is the biggest city in Iran's southeast, and they are the representative of this population. In the first phase of the KERCADRS, 250 postal codes (called seeds) were randomly selected from the list of city residents registered in the city post office. First, the location of each seed was determined on the map, and then, each of them was contacted. After identifying the number of people in each family, people over 15 years old were invited to participate in the study. If people do not answer despite calling twice, their neighbour was replacing them based on their seed address. In the first phase of the KERCADRS study, the collection of samples continued until 24 people was collected in each cluster. In the second phase of this study, people were called again, and 1008 people who met the criteria for entering our study were included.\nThe study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). All study participants gave informed consent. The method of data collection has already been published in the study of Najafipour et al.\n7\n\n\nSome risk factors, including gender, age, physical activity, smoking status, using tobacco products, body mass index (BMI), opium addiction and history of hypertension or diabetes, the duration of hypertension, and duration of diabetes disease, were reviewed in our study.\nPeople's height was measured in a standing position without shoes. All participants in the study were weighed with minimal clothing and without shoes on a digital scale (Seca, model 707, Germany). In order to calculate the body mass index (BMI) of the participants, the weight (kg) of the patients was divided by the square of their height (m2). Patients with a BMI between 25–29.9 Kg/m2 were considered overweight, and patients over 30 Kg/m2 were considered obese.\nAccording to the guidelines of the World Health Organization, patients' blood pressure was measured with a standard manometer (mercury manometer RICHTER, Germany) after a 10‐min rest, and if it was abnormal, it was measured again after 1 h.\nWHO Global Physical Activity Questionnaire (GPAQ) and Metabolic Equivalent (MET) were used to evaluate the amount and intensity of daily physical activity, respectively. A MET is defined as the amount of energy consumed while sitting (equivalent to 1 kcal/kg/h). Moderate to severe physical activity is defined as four times and eight times calorie intake compared to sitting quietly. The amount of activity more than 3000 MET and less than 1500 MET per week was considered intense and low physical activity, respectively.\n8\n\n\nA trained interviewer asked participants about smoking, tobacco use and opium addiction. People who routinely smoked cigarettes during data collection were considered smokers. Also, people who continuously used tobacco products or opium during data collection were considered positive for tobacco use and opium addiction, respectively.\nIn addition to determining demographic features of the studied people, being informed of this laboratory finding and drug treatment in the case of pre‐notice, a sample of fasting blood was taken (after 12–14 h of fasting) to determine the level of serum uric acid and fasting plasma glucose (FPG) for each person.\nFPG (KIMIA Kit, code 890410, Iran) and uric acid (measured by the colorimetric method.) were measured for all patients. In our study, the hyperuricaemia cut‐off was considered to be greater than the 75th percentile, which was 6 mg/dl in males and 5 mg/dl in females.\nThe senior researcher trained all the interviewees prior to data collection, and he also continuously evaluated the validity of the data collected. Although agreement indices were not measured in this study, careful quality control was performed to minimize errors.\nSPSS software version 16 (SPSS Inc.) was used to analyse data. Categorical variables were reported as numbers and percentages, and a chi‐square test was used to compare them between the two groups. Continuous variables were reported as medians (with interquartile ranges [IQRs]) based on their abnormal distribution according to the Kolmogorov–Smirnov test, and the Mann–Whitney test was used to compare them between individuals with and without hyperuricaemia. p < .05 was considered as significant.\n[SUBTITLE] Sample size estimation [SUBSECTION] To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\n9\n and considering Z = 1.96, p = .21, and D = 0.03:\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\n\n\nTo increase the power of our regression, we investigated 1008 people in our study.\nTo determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\n9\n and considering Z = 1.96, p = .21, and D = 0.03:\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\n\n\nTo increase the power of our regression, we investigated 1008 people in our study.\n[SUBTITLE] Statistical analysis [SUBSECTION] To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.\nTo determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.","To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\n9\n and considering Z = 1.96, p = .21, and D = 0.03:\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\n\n\nTo increase the power of our regression, we investigated 1008 people in our study.","To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.","A number of 1008 cases were entered into the study. The relationship between the studied variables and the prevalence of hyperuricaemia is shown in Table 1. According to the results of this study, 254 patients had uric acid levels above the 75th percentile (more than 6 mg/dl in males, and 5 mg/dl in females). No significant difference was seen between gender (p = .249) and age groups (p = .125) of people with and without hyperuricaemia. The prevalence of overweight/obesity was significantly higher in patients with hyperuricaemia (p < .001). The prevalence of low physical activity in patients with hyperuricaemia was significantly higher than in those with normal uric acid levels (p = .033). The prevalence of smoking (p = 406), tobacco consumption (p = .669) and opium addiction (p = .261) in people with and without hyperuricaemia did not show a significant difference. The frequency of hypertension was significantly higher in people with high uric acid levels compared to people with normal uric acid levels (31.9% and 22.8%, respectively; p = .004). The duration of hypertension was significantly higher in hyperuricaemic individuals (p = .022). The prevalence of diabetes (p = .794) and the duration of diabetes (p = .371) in people with and without hyperuricaemia did not show a significant difference. The duration of hypertension and diabetes was converted into categorized variables according to their median (48 and 72 months, respectively). The prevalence of a longer period of hypertension (more than 48 months) in people with hyperuricaemia was significantly higher than in those without hyperuricaemia (p = .025). The prevalence of a longer duration of diabetes (more than 72 months) between the two groups of people with and without hyperuricaemia was similar (p = .986).\nThe relationship between the studied variables and the presence of hyperuricaemia\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h)).\nChi‐square test.\nMann–Whitney test.\nIn order to determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used on variables whose frequency was significantly different in individuals with and without hyperuricaemia (Table 2). Based on the univariate logistic regression model, the presence of overweight/obesity, low physical activity and hypertension increased the chances of hyperuricaemia in the subjects by 2.85, 1.91, compared to those with intense physical activity and 1.58, respectively (p < .001, p = .019 and p = .004, respectively). According to the results of a multivariate analysis on the three variables, the amount of physical activity did not have a significant effect on increasing the chances of hyperuricaemia and the two variables of overweight/obesity (OR = 2.67; 95% CI = 1.87–3.82) and hypertension (OR = 1.40; 95% CI = 1.02–1.93) significantly increased the chances of hyperuricaemia in the subjects.\nUnivariate and multivariate logistic regression models for identifying determinants of hyperuricaemia\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h)).","According to the results of our study, overweight/obesity and hypertension increase the risk of hyperuricaemia in the general population. Due to the increasing prevalence of hyperuricaemia, it should be considered as an important health issue. So far, several studies have emphasized the relationship between the presence of hyperuricaemia and cardiovascular risk factors, including hypertension and metabolic syndrome\n10\n, \n11\n, \n12\n as in a recent cross‐sectional study; Yazdi et al.\n13\n showed that the ratio of uric acid to high‐density lipoprotein (HDL) cholesterol is significantly higher in people with metabolic syndrome than in healthy individuals. Despite the similarity of the results of the present study with previous studies, there are also differences. Different study designs and different criteria for confirming hyperuricaemia in patients can be the main cause of these differences.\nIn previous studies, inflammatory response, oxidative stress, insulin resistance, endothelial dysfunction and endoplasmic reticulum stress have been mentioned as the main mechanisms of hyperuricaemia's effect on the development of cardiovascular diseases.\n14\n In our study, hypertension was associated with increased uric acid levels in patients. In line with our study in Zhang et al.,\n15\n a prospective cohort study of 6424 patients, hypertension was also associated with an increased risk of hyperuricaemia. In another cross‐sectional study of 3119 people over the age of 50 in a rural population in China, high blood pressure increased the risk of high uric acid levels.\n10\n Although the mechanism of association between hyperuricaemia and hypertension is still unclear, according to the findings of in vitro studies, uric acid causes damage to endothelial cells, vasoconstriction and high blood pressure by inhibiting the release of nitric oxide from endothelial cells, activating the renin‐angiotensin system and oxidative stress.\n16\n Chen et al.’s\n17\n study of animal models and generally healthy early Parkinson's disease patients found no association between uric acid levels and blood pressure. Another possible cause of elevated uric acid in hypertensive patients is the use of certain drugs, including diuretics, β‐blockers, angiotensin‐converting enzyme inhibitors and non‐losartan angiotensin II receptor blockers.\n18\n In our study, the duration of hypertension in patients with hyperuricaemia was significantly longer, which could be due to the long‐term use of antihypertensive drugs, so more studies should be performed on new cases of hypertension to study the relationship between hypertension and uric acid levels.\nIn our study, overweight/obesity was associated with hyperuricaemia. In a study by Choi et al.,\n19\n BMI higher than 25 was significantly associated with an increased risk of gout in patients, and weight loss was a protective factor against gout. The link between obesity and hyperuricaemia can be due to insulin resistance caused by oxidative stress and inflammatory processes. Hyperinsulinaemia is associated with insulin resistance and increases the activity and expression of urate transporter protein 1 (URAT1) and, consequently, uric acid reabsorption through renal proximal tubular cells.\n14\n, \n20\n However, in our study, no association was found between hyperuricaemia and diabetes. In a study by Krishnan et al.,\n5\n elevated serum urate levels were an independent risk factor for type 2 diabetes in non‐obese patients. In a study by Van der Schaft et al.,\n6\n serum uric acid was associated with incidental prediabetes in patients with normal blood sugar but not with the development of type 2 diabetes in prediabetic subjects. In a study conducted on mice by Lu et al.,\n21\n there was no association between high uric acid levels and diabetes, and they concluded that high uric acid levels did not cause diabetes but could accelerate it by damaging glucose tolerance.\nIn our study, although less physical activity significantly increased the risk of hyperuricaemia in univariate analysis, the relationship between physical activity and hyperuricaemia was not significant in multivariate analysis after adjustment of the effect of overweight/obesity and hypertension variables. Dong et al.,\n22\n in a study of 38,855 rural China populations, showed that regular physical activity independently reduced serum uric acid levels and the prevalence of hyperuricaemia. However, in the regression analysis performed in the study of Dong et al.,\n22\n adjustment was done for BMI but not for hypertension. In the study by Qi et al.,\n10\n low physical activity in men was associated with an increased risk of high uric acid levels. It seems that racial differences in the relationship between physical activity and hyperuricaemia need to be further evaluated.\nIn this study, the prevalence of hyperuricaemia has no significant relationship with age groups, smoking, tobacco consumption and opium addiction. Despite our study, Hanna et al.\n23\n showed that uric acid serum level is significantly lower in smoking people, but these people did not suffer from hyperuricaemia in our study. It was shown in another study conducted by Tomita et al.\n24\n that, like our study, uric acid serum levels and hyperuricaemia in smoking people have no significant difference with people who have never smoked statistically. In this study, the majority of cases belonged to the age group of 41 to 60 (45.1%) years, which probably people aged 41–60 had a higher tendency to participate in the study due to their awareness of the benefits of the study.\nOne of the strengths of our study was its population‐based and appropriate sample size. The sampling method in our study was cluster. Among the advantages of this method is that it is economical in terms of cost and time compared to probability sampling methods. But the accuracy of this method is lower than simple random sampling. Our study also had limitations, including the fact that many factors, including smoking and the duration of hypertension, were self‐reported, and there was a possibility of bias in these variables. Another limitation of the study was the possible effects of the drugs used and their lack of evaluation, including diuretics, aspirin and dietary habits on patients' uric acid levels separately. Another limitation of our study was that, unfortunately, the dietary profile of the participants was not recorded at the time of sampling. Despite all the limitations mentioned, our findings are helpful in determining the appropriate strategy to identify people who are most at risk for hyperuricaemia for the prevention and early treatment of this disorder.","Our study showed that two independent factors for the development of hyperuricaemia are hypertension and overweight/obesity. Since hyperuricaemia is one of the influential factors in cardiovascular events, these events can be prevented by determining the appropriate strategy for early diagnosis and treatment of this metabolic disorder.\nConsidering that our study was conducted in an urban population and environmental and genetic factors can also be involved in the relationship between hyperuricaemia and other cardiovascular disease risk factors, cohort studies with a larger sample size are needed in this field.","Farzaneh Yazdi, Mohammad Reza Shakibi, Hamid Najafipour, Amir Baniasad, Mohammad Javad Najafzadeh and Samira Sistani designed the study and collected data, Amir Baniasad, Mohammad Javad Najafzadeh wrote the first draft of the paper and submitted the manuscript, Fatemeh Yazdi contributed to writing and revision of the manuscript. All authors contributed to finalizing the manuscript.","This research project was funded by the Kerman University of Medical Sciences.","The authors declare that they have no conflict of interest.","The study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). Informed consent was obtained from all participants in the study."],"string":"[\n \"Cardiovascular diseases are one of the most important causes of death in the world, and their pathogenesis is such that they are affected by several risk factors such as obesity, diabetes and hypertension.\\n1\\n Increased serum uric acid levels can lead to endothelial dysfunction resulting in atherosclerosis by releasing reactive oxygen species and is reported as one of the risk factors for cardiovascular diseases.\\n2\\n\\n\\nIn the process of purine decomposition, uric acid is the final product.\\n3\\n A condition in which serum uric acid concentration is high but there are no signs or symptoms of uric acid deposition is called asymptomatic hyperuricaemia. Although gout can occur in a person with any rate of uric acid, about two‐thirds of hyperuricaemic people remain asymptomatic.\\n4\\n Different studies in different races have had different results on the prevalence of hyperuricaemia in people's age and gender.\\n5\\n, \\n6\\n\\n\\nPrevalence of hyperuricaemia is different in viewpoints of race, age and gender groups, and this serum factor has been mentioned as a risk factor for coronary artery diseases; in this study, the relationship between cardiovascular risk factors and hyperuricaemia was investigated.\",\n \"This study was a sub‐analyse of collected data about the risk factors of coronary artery diseases (i.e., Kerman Coronary Artery Disease Risk Factor Study; KERCADRS). This cross‐sectional and descriptive study was conducted in Kerman city on 1008 people who were over 15 years old. The sampling method of the studied people was in the form of a cluster from the entire population of the city. These people were selected from 720,000 Kerman residents, which is the biggest city in Iran's southeast, and they are the representative of this population. In the first phase of the KERCADRS, 250 postal codes (called seeds) were randomly selected from the list of city residents registered in the city post office. First, the location of each seed was determined on the map, and then, each of them was contacted. After identifying the number of people in each family, people over 15 years old were invited to participate in the study. If people do not answer despite calling twice, their neighbour was replacing them based on their seed address. In the first phase of the KERCADRS study, the collection of samples continued until 24 people was collected in each cluster. In the second phase of this study, people were called again, and 1008 people who met the criteria for entering our study were included.\\nThe study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). All study participants gave informed consent. The method of data collection has already been published in the study of Najafipour et al.\\n7\\n\\n\\nSome risk factors, including gender, age, physical activity, smoking status, using tobacco products, body mass index (BMI), opium addiction and history of hypertension or diabetes, the duration of hypertension, and duration of diabetes disease, were reviewed in our study.\\nPeople's height was measured in a standing position without shoes. All participants in the study were weighed with minimal clothing and without shoes on a digital scale (Seca, model 707, Germany). In order to calculate the body mass index (BMI) of the participants, the weight (kg) of the patients was divided by the square of their height (m2). Patients with a BMI between 25–29.9 Kg/m2 were considered overweight, and patients over 30 Kg/m2 were considered obese.\\nAccording to the guidelines of the World Health Organization, patients' blood pressure was measured with a standard manometer (mercury manometer RICHTER, Germany) after a 10‐min rest, and if it was abnormal, it was measured again after 1 h.\\nWHO Global Physical Activity Questionnaire (GPAQ) and Metabolic Equivalent (MET) were used to evaluate the amount and intensity of daily physical activity, respectively. A MET is defined as the amount of energy consumed while sitting (equivalent to 1 kcal/kg/h). Moderate to severe physical activity is defined as four times and eight times calorie intake compared to sitting quietly. The amount of activity more than 3000 MET and less than 1500 MET per week was considered intense and low physical activity, respectively.\\n8\\n\\n\\nA trained interviewer asked participants about smoking, tobacco use and opium addiction. People who routinely smoked cigarettes during data collection were considered smokers. Also, people who continuously used tobacco products or opium during data collection were considered positive for tobacco use and opium addiction, respectively.\\nIn addition to determining demographic features of the studied people, being informed of this laboratory finding and drug treatment in the case of pre‐notice, a sample of fasting blood was taken (after 12–14 h of fasting) to determine the level of serum uric acid and fasting plasma glucose (FPG) for each person.\\nFPG (KIMIA Kit, code 890410, Iran) and uric acid (measured by the colorimetric method.) were measured for all patients. In our study, the hyperuricaemia cut‐off was considered to be greater than the 75th percentile, which was 6 mg/dl in males and 5 mg/dl in females.\\nThe senior researcher trained all the interviewees prior to data collection, and he also continuously evaluated the validity of the data collected. Although agreement indices were not measured in this study, careful quality control was performed to minimize errors.\\nSPSS software version 16 (SPSS Inc.) was used to analyse data. Categorical variables were reported as numbers and percentages, and a chi‐square test was used to compare them between the two groups. Continuous variables were reported as medians (with interquartile ranges [IQRs]) based on their abnormal distribution according to the Kolmogorov–Smirnov test, and the Mann–Whitney test was used to compare them between individuals with and without hyperuricaemia. p < .05 was considered as significant.\\n[SUBTITLE] Sample size estimation [SUBSECTION] To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\\n9\\n and considering Z = 1.96, p = .21, and D = 0.03:\\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\\n\\n\\nTo increase the power of our regression, we investigated 1008 people in our study.\\nTo determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\\n9\\n and considering Z = 1.96, p = .21, and D = 0.03:\\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\\n\\n\\nTo increase the power of our regression, we investigated 1008 people in our study.\\n[SUBTITLE] Statistical analysis [SUBSECTION] To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.\\nTo determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.\",\n \"To determine the sample size based on the study of Zhu et al., the prevalence of hyperuricaemia was considered to be 21.4%, and the minimum required sample size was calculated using the following formula\\n9\\n and considering Z = 1.96, p = .21, and D = 0.03:\\nn=Z2P1−Pd2=1.9620.211−0.210.032=708\\n\\n\\nTo increase the power of our regression, we investigated 1008 people in our study.\",\n \"To determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used. Variables significantly different in individuals with and without hyperuricaemia were included in the regression models, and the odds ratios (AOR) were reported.\",\n \"A number of 1008 cases were entered into the study. The relationship between the studied variables and the prevalence of hyperuricaemia is shown in Table 1. According to the results of this study, 254 patients had uric acid levels above the 75th percentile (more than 6 mg/dl in males, and 5 mg/dl in females). No significant difference was seen between gender (p = .249) and age groups (p = .125) of people with and without hyperuricaemia. The prevalence of overweight/obesity was significantly higher in patients with hyperuricaemia (p < .001). The prevalence of low physical activity in patients with hyperuricaemia was significantly higher than in those with normal uric acid levels (p = .033). The prevalence of smoking (p = 406), tobacco consumption (p = .669) and opium addiction (p = .261) in people with and without hyperuricaemia did not show a significant difference. The frequency of hypertension was significantly higher in people with high uric acid levels compared to people with normal uric acid levels (31.9% and 22.8%, respectively; p = .004). The duration of hypertension was significantly higher in hyperuricaemic individuals (p = .022). The prevalence of diabetes (p = .794) and the duration of diabetes (p = .371) in people with and without hyperuricaemia did not show a significant difference. The duration of hypertension and diabetes was converted into categorized variables according to their median (48 and 72 months, respectively). The prevalence of a longer period of hypertension (more than 48 months) in people with hyperuricaemia was significantly higher than in those without hyperuricaemia (p = .025). The prevalence of a longer duration of diabetes (more than 72 months) between the two groups of people with and without hyperuricaemia was similar (p = .986).\\nThe relationship between the studied variables and the presence of hyperuricaemia\\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h)).\\nChi‐square test.\\nMann–Whitney test.\\nIn order to determine the potential predictors of hyperuricaemia, univariable and multivariable survey logistic regression models were used on variables whose frequency was significantly different in individuals with and without hyperuricaemia (Table 2). Based on the univariate logistic regression model, the presence of overweight/obesity, low physical activity and hypertension increased the chances of hyperuricaemia in the subjects by 2.85, 1.91, compared to those with intense physical activity and 1.58, respectively (p < .001, p = .019 and p = .004, respectively). According to the results of a multivariate analysis on the three variables, the amount of physical activity did not have a significant effect on increasing the chances of hyperuricaemia and the two variables of overweight/obesity (OR = 2.67; 95% CI = 1.87–3.82) and hypertension (OR = 1.40; 95% CI = 1.02–1.93) significantly increased the chances of hyperuricaemia in the subjects.\\nUnivariate and multivariate logistic regression models for identifying determinants of hyperuricaemia\\nAbbreviations: BMI: Body mass index, MET: Metabolic Equivalent (The amount of energy consumed while sitting (equivalent to 1 kcal/kg/h)).\",\n \"According to the results of our study, overweight/obesity and hypertension increase the risk of hyperuricaemia in the general population. Due to the increasing prevalence of hyperuricaemia, it should be considered as an important health issue. So far, several studies have emphasized the relationship between the presence of hyperuricaemia and cardiovascular risk factors, including hypertension and metabolic syndrome\\n10\\n, \\n11\\n, \\n12\\n as in a recent cross‐sectional study; Yazdi et al.\\n13\\n showed that the ratio of uric acid to high‐density lipoprotein (HDL) cholesterol is significantly higher in people with metabolic syndrome than in healthy individuals. Despite the similarity of the results of the present study with previous studies, there are also differences. Different study designs and different criteria for confirming hyperuricaemia in patients can be the main cause of these differences.\\nIn previous studies, inflammatory response, oxidative stress, insulin resistance, endothelial dysfunction and endoplasmic reticulum stress have been mentioned as the main mechanisms of hyperuricaemia's effect on the development of cardiovascular diseases.\\n14\\n In our study, hypertension was associated with increased uric acid levels in patients. In line with our study in Zhang et al.,\\n15\\n a prospective cohort study of 6424 patients, hypertension was also associated with an increased risk of hyperuricaemia. In another cross‐sectional study of 3119 people over the age of 50 in a rural population in China, high blood pressure increased the risk of high uric acid levels.\\n10\\n Although the mechanism of association between hyperuricaemia and hypertension is still unclear, according to the findings of in vitro studies, uric acid causes damage to endothelial cells, vasoconstriction and high blood pressure by inhibiting the release of nitric oxide from endothelial cells, activating the renin‐angiotensin system and oxidative stress.\\n16\\n Chen et al.’s\\n17\\n study of animal models and generally healthy early Parkinson's disease patients found no association between uric acid levels and blood pressure. Another possible cause of elevated uric acid in hypertensive patients is the use of certain drugs, including diuretics, β‐blockers, angiotensin‐converting enzyme inhibitors and non‐losartan angiotensin II receptor blockers.\\n18\\n In our study, the duration of hypertension in patients with hyperuricaemia was significantly longer, which could be due to the long‐term use of antihypertensive drugs, so more studies should be performed on new cases of hypertension to study the relationship between hypertension and uric acid levels.\\nIn our study, overweight/obesity was associated with hyperuricaemia. In a study by Choi et al.,\\n19\\n BMI higher than 25 was significantly associated with an increased risk of gout in patients, and weight loss was a protective factor against gout. The link between obesity and hyperuricaemia can be due to insulin resistance caused by oxidative stress and inflammatory processes. Hyperinsulinaemia is associated with insulin resistance and increases the activity and expression of urate transporter protein 1 (URAT1) and, consequently, uric acid reabsorption through renal proximal tubular cells.\\n14\\n, \\n20\\n However, in our study, no association was found between hyperuricaemia and diabetes. In a study by Krishnan et al.,\\n5\\n elevated serum urate levels were an independent risk factor for type 2 diabetes in non‐obese patients. In a study by Van der Schaft et al.,\\n6\\n serum uric acid was associated with incidental prediabetes in patients with normal blood sugar but not with the development of type 2 diabetes in prediabetic subjects. In a study conducted on mice by Lu et al.,\\n21\\n there was no association between high uric acid levels and diabetes, and they concluded that high uric acid levels did not cause diabetes but could accelerate it by damaging glucose tolerance.\\nIn our study, although less physical activity significantly increased the risk of hyperuricaemia in univariate analysis, the relationship between physical activity and hyperuricaemia was not significant in multivariate analysis after adjustment of the effect of overweight/obesity and hypertension variables. Dong et al.,\\n22\\n in a study of 38,855 rural China populations, showed that regular physical activity independently reduced serum uric acid levels and the prevalence of hyperuricaemia. However, in the regression analysis performed in the study of Dong et al.,\\n22\\n adjustment was done for BMI but not for hypertension. In the study by Qi et al.,\\n10\\n low physical activity in men was associated with an increased risk of high uric acid levels. It seems that racial differences in the relationship between physical activity and hyperuricaemia need to be further evaluated.\\nIn this study, the prevalence of hyperuricaemia has no significant relationship with age groups, smoking, tobacco consumption and opium addiction. Despite our study, Hanna et al.\\n23\\n showed that uric acid serum level is significantly lower in smoking people, but these people did not suffer from hyperuricaemia in our study. It was shown in another study conducted by Tomita et al.\\n24\\n that, like our study, uric acid serum levels and hyperuricaemia in smoking people have no significant difference with people who have never smoked statistically. In this study, the majority of cases belonged to the age group of 41 to 60 (45.1%) years, which probably people aged 41–60 had a higher tendency to participate in the study due to their awareness of the benefits of the study.\\nOne of the strengths of our study was its population‐based and appropriate sample size. The sampling method in our study was cluster. Among the advantages of this method is that it is economical in terms of cost and time compared to probability sampling methods. But the accuracy of this method is lower than simple random sampling. Our study also had limitations, including the fact that many factors, including smoking and the duration of hypertension, were self‐reported, and there was a possibility of bias in these variables. Another limitation of the study was the possible effects of the drugs used and their lack of evaluation, including diuretics, aspirin and dietary habits on patients' uric acid levels separately. Another limitation of our study was that, unfortunately, the dietary profile of the participants was not recorded at the time of sampling. Despite all the limitations mentioned, our findings are helpful in determining the appropriate strategy to identify people who are most at risk for hyperuricaemia for the prevention and early treatment of this disorder.\",\n \"Our study showed that two independent factors for the development of hyperuricaemia are hypertension and overweight/obesity. Since hyperuricaemia is one of the influential factors in cardiovascular events, these events can be prevented by determining the appropriate strategy for early diagnosis and treatment of this metabolic disorder.\\nConsidering that our study was conducted in an urban population and environmental and genetic factors can also be involved in the relationship between hyperuricaemia and other cardiovascular disease risk factors, cohort studies with a larger sample size are needed in this field.\",\n \"Farzaneh Yazdi, Mohammad Reza Shakibi, Hamid Najafipour, Amir Baniasad, Mohammad Javad Najafzadeh and Samira Sistani designed the study and collected data, Amir Baniasad, Mohammad Javad Najafzadeh wrote the first draft of the paper and submitted the manuscript, Fatemeh Yazdi contributed to writing and revision of the manuscript. All authors contributed to finalizing the manuscript.\",\n \"This research project was funded by the Kerman University of Medical Sciences.\",\n \"The authors declare that they have no conflict of interest.\",\n \"The study protocol was reviewed and approved by the ethics committee of Kerman University of Medical Sciences (ethic code: IR.KMU.REC.1394.59). Informed consent was obtained from all participants in the study.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"materials-and-methods",null,null,"results","discussion","conclusions",null,null,"COI-statement",null],"string":"[\n null,\n \"materials-and-methods\",\n null,\n null,\n \"results\",\n \"discussion\",\n \"conclusions\",\n null,\n null,\n \"COI-statement\",\n null\n]"},"keywords":{"kind":"list like","value":["diabetes","hypertension","hyperuricaemia","obesity","overweight","physical activity"],"string":"[\n \"diabetes\",\n \"hypertension\",\n \"hyperuricaemia\",\n \"obesity\",\n \"overweight\",\n \"physical activity\"\n]"}}},{"rowIdx":378,"cells":{"title":{"kind":"string","value":"Second Malignant Tumors and Non-Tumor Causes of Death for Patients With Penile Cancer During Their Survivorship."},"pmid":{"kind":"string","value":"36267038"},"background_abstract":{"kind":"string","value":"The aim was to evaluate the causes of death for patients with localized, regional and metastatic penile cancer (PeCa) after diagnosis."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"PeCa patients diagnosed during 2004-2018 in the Surveillance, Epidemiology, and End Results program database were identified. Causes of deaths including PeCa, second malignant tumors (SMTs) and non-tumor diseases were analyzed, as well as the standardized mortality ratio (SMR) of each cause."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"For localized PeCa, 800 of 2155 patients died during the follow-up. 24.9% of all deaths were due to PeCa. 18.0% and 57.1% deaths were due to SMTs and non-tumor causes. Main SMTs included cancers of lung and bronchus (n = 40) and skin (n = 11) with significantly increased SMRs of 1.71 (1.22-2.33) and 4.82 (2.41-8.63). Mortality risks of other SMTs were mostly similar with the general populations. Main causes of non-tumor diseases included diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], COPD and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], and cerebrovascular diseases [n = 33, SMR: 1.71 (1.17-2.4)]. For regional PeCa, 679 of 1310 patients died including 43.5% PeCa, 14.8% SMTs and 26.6% non-tumor causes. The mortality risks of cancers from lung and bronchus [SMR: 2.41 (1.53-3.62)], skin [SMR: 6.41 (2.35-13.95)] and testis [SMR: 149.35 (18.09-539.5)] were significantly increased. Main non-tumor causes of death included diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)] and diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)]. For distant diseases, 109 of 132 patients died including 76 (69.7%) died for PeCa itself, 24 (22.0%) died for SMTs and 9 (8.3%) died for non-tumor diseases. The majority of PeCa deaths (67.1%) and SMTs deaths (79.2%) occurred within 1 year after the diagnosis of PeCa."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"We firstly analyzed the SMTs and non-tumor causes of death and morality risks of each cause for PeCa patients, which provided valuable information for PeCa patients on disease prevention and health care during their survivorship."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Male","Humans","Cause of Death","Penile Neoplasms","Survivorship","Risk Factors","Pulmonary Disease, Chronic Obstructive"],"string":"[\n \"Male\",\n \"Humans\",\n \"Cause of Death\",\n \"Penile Neoplasms\",\n \"Survivorship\",\n \"Risk Factors\",\n \"Pulmonary Disease, Chronic Obstructive\"\n]"},"pmcid":{"kind":"string","value":"9597479"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Penile cancer (PeCa) is a rare malignancy worldwide, accounting for less than 1% of all malignancies in men.1 The overall incidence rate is approximately .69 cases per 100 000 persons in the US.2 It was estimated that the number of new PeCa patients was 34 475 and deaths cases was 15 138 each year worldwide.3 The survival rates still vary greatly with the stages of patients’ disease. It was reported that about 40% of new cases were diagnosed with localized PeCa, with a 5-year overall survival rate of approximately 90%.4 The survival rates decrease dramatically once patients are in the regional or distant metastasis stage. The 5-year overall survival rate is approximately 80% with unilateral inguinal lymph node involvement, 10%-20% with bilateral or pelvic lymph node involvement, and <10% with extranodal extension.4\nSince the 1990s, substantial progress has been made in the prevention, diagnosis and treatment of PeCa. The prognosis has been greatly improved in PeCa patients, especially those who have not progressed to the metastatic stage. With the prolonged survival time, the proportion of other causes of death such as second malignant tumors (SMTs) and non-tumor diseases has been increasing among all deaths of PeCa patients. Since few previous studies have focused exclusively on other causes of death of PeCa patients, an analysis of causes of death among PeCa patients is warranted. Our study aimed to evaluate the main causes of death including SMTs and non-cancer causes for PeCa patients, and calculate the standardized mortality ratio (SMR) for each reason of death compared with the normal population."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Data Sources [SUBSECTION] The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\nThe data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\n[SUBTITLE] Study Population and Study Variables [SUBSECTION] Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\nPatients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\n[SUBTITLE] Outcome Assessments [SUBSECTION] The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\nThe causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\n[SUBTITLE] Statistics Analyses [SUBSECTION] For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\nFor the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"A total of 3597 PeCa patients were included, 1588 (44%) of them died during the follow-up. The majority of patients was in the localized stage (n = 2,155, 60%) and regional stage (n = 1310, 36%), whereas only 4% (n = 132) had distant disease. For patients with localized PeCa, 800 (37.1%) of them died with a SMR of 2.12 (1.97-2.27). 679 (51.8%) regional PeCa patients [SMR: 4.61 (4.27-4.97)] and 109 (82.6%) patients with distant metastatic PeCa [SMR: 27.68 (22.73-33.39)] died during the follow-up.\nBaseline characteristics of included patients and the dead cases were presented in Table 1. The trend of new diagnosed PeCa patients from 2000 to 2018 were revealed in Figure 1.Table 1.Baseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.CharacteristicLocalizedRegionalDistantPatients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Total21558002.12a (1.97-2.27)13106794.61a (4.27-4.97)13210927.68a (22.73-33.39)Age 15-54 years470804.41a (3.5-5.49)32912116.05a (13.32-19.18)3630106.43a (71.81-151.93) 55-64 years4991282.86a (2.39-3.4)3371537.48a (6.34-8.77)322748.63a (32.05-70.75) 65-74 years5652102.21a (1.92-2.53)3301834.67a (4.01-5.39)332635.53a (23.21-52.06) 75-84 years4102221.55a (1.35-1.76)2171432.61a (2.2-3.08)262110.51a (6.51-16.07) 85+ years2111602.10a (1.79-2.45)97793.10a (2.45-3.86)5513.45a (4.37-31.39)Year of diagnosis 2004-20087301432.65a (2.23-3.12)3532343.11a (2.73-3.54)302717.20a (11.34-25.03) 2009-20137442842.05a (1.82-2.3)4422585.32a (4.69-6.01)484232.68a (23.55-44.17) 2014-20186813732.01a (1.81-2.23)5151877.88a (6.79-9.1)544036.91a (26.37-50.27)Race White18066782.08a (1.92-2.24)11205774.35a (4-4.72)1058326.90a (21.43-33.35) Black222842.48a(1.98-3.07)116727.28a (5.7-9.17)161520.92a (11.71-34.51) American Indian/Alaska native2183.79a (1.64-7.47)1137.25a (1.49-21.18)33586.91a (121.03-1715.19) Asian or Pacific islander106301.95a (1.31-2.78)63276.22a (4.1-9.05)8861.02a (26.35-120.24)Grade Well differentiated6312181.78a (1.55-2.03)188863.01a (2.41-3.71)7620.27a (7.44-44.12) Moderately differentiated7063042.55a (2.27-2.85)5743175.07a (4.52-5.65)383219.92a (13.62-28.11) Poorly differentiated2571302.93a (2.44-3.47)3171925.02a (4.34-5.79)444135.67a (25.6-48.39) Undifferentiated92.7 (.08-2.52)1175.63a (2.27-11.61)329.89a (1.2-35.72)Pathological type 8070/3: Squamous cell carcinoma, NOS14255602.22a (2.04-2.41)8294474.49a (4.08-4.93)846830.34a (23.56-38.46) 8071/3: Squamous cell carcinoma, keratinizing, NOS3471262.26a (1.88-2.69)3321655.54a (4.73-6.46)191414.96a (8.18-25.1) 8051/3: Verrucous carcinoma, NOS169491.50a (1.11-1.98)39101.54 (.74-2.84)2298.29a (11.9-355.06) 8083/3: Basaloid squamous cell carcinoma41122.96a (1.53-5.18)34165.50a(3.14-8.93)6523.04a (7.48-53.77)Radiotherapy Beam radiation68301.90a (1.28-2.72)1981288.26a(6.89-9.83)372945.93a(30.76-65.96) Radioactive implants/brachytherapy214.79 (.12-26.7)3112.38 (.31-68.98)000 None/unknown20857692.13a (1.98-2.28)11095504.17a(3.83-4.54)958024.19a(19.18-30.11)Chemotherapy Yes59222.69a (1.69-4.08)29816711.38a(9.72-13.24)745831.87a(24.2-41.21) No/unknown20967782.10a (1.96-2.26)10125123.86a(3.53-4.21)585124.07a(17.92-31.65)SMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.aStatistical significance with P < .05.Figure 1.The number of newly diagnosed of penile cancer from 2004 to 2018.\nBaseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.\nSMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.\naStatistical significance with P < .05.\nThe number of newly diagnosed of penile cancer from 2004 to 2018.\n[SUBTITLE] Causes of Death for Patients With Localized Penile Cancer [SUBSECTION] Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nAmong the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\n[SUBTITLE] Causes of Death for Patients With Regional Penile Cancer [SUBSECTION] For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.\nFor patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.\n[SUBTITLE] Causes of Death for Patients With Distant Metastatic Penile Cancer [SUBSECTION] For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05.\nFor the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05."},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"Our study provided a detailed analysis of the causes of death for patients with locally, regionally, and distant metastatic PeCa after diagnosis. This information could be useful for disease prevention and health care during patients’ survivorship."},"other_sections_titles":{"kind":"list like","value":["Data Sources","Study Population and Study Variables","Outcome Assessments","Statistics Analyses","Causes of Death for Patients With Localized Penile Cancer","Causes of Death for Patients With Regional Penile Cancer","Causes of Death for Patients With Distant Metastatic Penile Cancer"],"string":"[\n \"Data Sources\",\n \"Study Population and Study Variables\",\n \"Outcome Assessments\",\n \"Statistics Analyses\",\n \"Causes of Death for Patients With Localized Penile Cancer\",\n \"Causes of Death for Patients With Regional Penile Cancer\",\n \"Causes of Death for Patients With Distant Metastatic Penile Cancer\"\n]"},"other_sections_texts":{"kind":"list like","value":["The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.","Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).","The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.","For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.","Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.","For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.","For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05."],"string":"[\n \"The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\",\n \"Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\",\n \"The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\",\n \"For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\",\n \"Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nMain causes of death for patients with localized penile cancer.\\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\\naStatistical significance with P < .05.\\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\",\n \"For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\\nMain causes of death for patients with regional penile cancer.\\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\\naStatistical significance with P < .05.\",\n \"For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\\nMain Causes of Death for Patients With Distant Penile Cancer.\\nSMR: standardized mortality ratio; CI: confidence interval.\\naStatistical significance with P < .05.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Data Sources","Study Population and Study Variables","Outcome Assessments","Statistics Analyses","Results","Causes of Death for Patients With Localized Penile Cancer","Causes of Death for Patients With Regional Penile Cancer","Causes of Death for Patients With Distant Metastatic Penile Cancer","Discussion","Conclusion","Supplemental Material"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Data Sources\",\n \"Study Population and Study Variables\",\n \"Outcome Assessments\",\n \"Statistics Analyses\",\n \"Results\",\n \"Causes of Death for Patients With Localized Penile Cancer\",\n \"Causes of Death for Patients With Regional Penile Cancer\",\n \"Causes of Death for Patients With Distant Metastatic Penile Cancer\",\n \"Discussion\",\n \"Conclusion\",\n \"Supplemental Material\"\n]"},"all_sections_texts":{"kind":"list like","value":["Penile cancer (PeCa) is a rare malignancy worldwide, accounting for less than 1% of all malignancies in men.1 The overall incidence rate is approximately .69 cases per 100 000 persons in the US.2 It was estimated that the number of new PeCa patients was 34 475 and deaths cases was 15 138 each year worldwide.3 The survival rates still vary greatly with the stages of patients’ disease. It was reported that about 40% of new cases were diagnosed with localized PeCa, with a 5-year overall survival rate of approximately 90%.4 The survival rates decrease dramatically once patients are in the regional or distant metastasis stage. The 5-year overall survival rate is approximately 80% with unilateral inguinal lymph node involvement, 10%-20% with bilateral or pelvic lymph node involvement, and <10% with extranodal extension.4\nSince the 1990s, substantial progress has been made in the prevention, diagnosis and treatment of PeCa. The prognosis has been greatly improved in PeCa patients, especially those who have not progressed to the metastatic stage. With the prolonged survival time, the proportion of other causes of death such as second malignant tumors (SMTs) and non-tumor diseases has been increasing among all deaths of PeCa patients. Since few previous studies have focused exclusively on other causes of death of PeCa patients, an analysis of causes of death among PeCa patients is warranted. Our study aimed to evaluate the main causes of death including SMTs and non-cancer causes for PeCa patients, and calculate the standardized mortality ratio (SMR) for each reason of death compared with the normal population.","[SUBTITLE] Data Sources [SUBSECTION] The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\nThe data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\n[SUBTITLE] Study Population and Study Variables [SUBSECTION] Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\nPatients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\n[SUBTITLE] Outcome Assessments [SUBSECTION] The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\nThe causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\n[SUBTITLE] Statistics Analyses [SUBSECTION] For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\nFor the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.","The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.","Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).","The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.","For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.","A total of 3597 PeCa patients were included, 1588 (44%) of them died during the follow-up. The majority of patients was in the localized stage (n = 2,155, 60%) and regional stage (n = 1310, 36%), whereas only 4% (n = 132) had distant disease. For patients with localized PeCa, 800 (37.1%) of them died with a SMR of 2.12 (1.97-2.27). 679 (51.8%) regional PeCa patients [SMR: 4.61 (4.27-4.97)] and 109 (82.6%) patients with distant metastatic PeCa [SMR: 27.68 (22.73-33.39)] died during the follow-up.\nBaseline characteristics of included patients and the dead cases were presented in Table 1. The trend of new diagnosed PeCa patients from 2000 to 2018 were revealed in Figure 1.Table 1.Baseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.CharacteristicLocalizedRegionalDistantPatients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Total21558002.12a (1.97-2.27)13106794.61a (4.27-4.97)13210927.68a (22.73-33.39)Age 15-54 years470804.41a (3.5-5.49)32912116.05a (13.32-19.18)3630106.43a (71.81-151.93) 55-64 years4991282.86a (2.39-3.4)3371537.48a (6.34-8.77)322748.63a (32.05-70.75) 65-74 years5652102.21a (1.92-2.53)3301834.67a (4.01-5.39)332635.53a (23.21-52.06) 75-84 years4102221.55a (1.35-1.76)2171432.61a (2.2-3.08)262110.51a (6.51-16.07) 85+ years2111602.10a (1.79-2.45)97793.10a (2.45-3.86)5513.45a (4.37-31.39)Year of diagnosis 2004-20087301432.65a (2.23-3.12)3532343.11a (2.73-3.54)302717.20a (11.34-25.03) 2009-20137442842.05a (1.82-2.3)4422585.32a (4.69-6.01)484232.68a (23.55-44.17) 2014-20186813732.01a (1.81-2.23)5151877.88a (6.79-9.1)544036.91a (26.37-50.27)Race White18066782.08a (1.92-2.24)11205774.35a (4-4.72)1058326.90a (21.43-33.35) Black222842.48a(1.98-3.07)116727.28a (5.7-9.17)161520.92a (11.71-34.51) American Indian/Alaska native2183.79a (1.64-7.47)1137.25a (1.49-21.18)33586.91a (121.03-1715.19) Asian or Pacific islander106301.95a (1.31-2.78)63276.22a (4.1-9.05)8861.02a (26.35-120.24)Grade Well differentiated6312181.78a (1.55-2.03)188863.01a (2.41-3.71)7620.27a (7.44-44.12) Moderately differentiated7063042.55a (2.27-2.85)5743175.07a (4.52-5.65)383219.92a (13.62-28.11) Poorly differentiated2571302.93a (2.44-3.47)3171925.02a (4.34-5.79)444135.67a (25.6-48.39) Undifferentiated92.7 (.08-2.52)1175.63a (2.27-11.61)329.89a (1.2-35.72)Pathological type 8070/3: Squamous cell carcinoma, NOS14255602.22a (2.04-2.41)8294474.49a (4.08-4.93)846830.34a (23.56-38.46) 8071/3: Squamous cell carcinoma, keratinizing, NOS3471262.26a (1.88-2.69)3321655.54a (4.73-6.46)191414.96a (8.18-25.1) 8051/3: Verrucous carcinoma, NOS169491.50a (1.11-1.98)39101.54 (.74-2.84)2298.29a (11.9-355.06) 8083/3: Basaloid squamous cell carcinoma41122.96a (1.53-5.18)34165.50a(3.14-8.93)6523.04a (7.48-53.77)Radiotherapy Beam radiation68301.90a (1.28-2.72)1981288.26a(6.89-9.83)372945.93a(30.76-65.96) Radioactive implants/brachytherapy214.79 (.12-26.7)3112.38 (.31-68.98)000 None/unknown20857692.13a (1.98-2.28)11095504.17a(3.83-4.54)958024.19a(19.18-30.11)Chemotherapy Yes59222.69a (1.69-4.08)29816711.38a(9.72-13.24)745831.87a(24.2-41.21) No/unknown20967782.10a (1.96-2.26)10125123.86a(3.53-4.21)585124.07a(17.92-31.65)SMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.aStatistical significance with P < .05.Figure 1.The number of newly diagnosed of penile cancer from 2004 to 2018.\nBaseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.\nSMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.\naStatistical significance with P < .05.\nThe number of newly diagnosed of penile cancer from 2004 to 2018.\n[SUBTITLE] Causes of Death for Patients With Localized Penile Cancer [SUBSECTION] Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nAmong the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\n[SUBTITLE] Causes of Death for Patients With Regional Penile Cancer [SUBSECTION] For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.\nFor patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.\n[SUBTITLE] Causes of Death for Patients With Distant Metastatic Penile Cancer [SUBSECTION] For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05.\nFor the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05.","Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nMain causes of death for patients with localized penile cancer.\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\naStatistical significance with P < .05.\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.","For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\nMain causes of death for patients with regional penile cancer.\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\naStatistical significance with P < .05.","For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\nMain Causes of Death for Patients With Distant Penile Cancer.\nSMR: standardized mortality ratio; CI: confidence interval.\naStatistical significance with P < .05.","With the high and increasing survival rate of PeCa, deaths from other causes account for a very high proportion of total deaths and tend to increase.5 Therefore, other causes of death including SMTs and non-tumor diseases should be taken seriously. Besides, whether these causes have relationships with the diagnosis of PeCa is unclear. In current literature, studies on the SMTs for the primary PeCa were quite rare. Our study, as far as we can know, is the first study focusing on the SMTs and non-cancer causes of death for primary PeCa patients. We also relate these causes to the time period after diagnosis of penile cancer and provide results stratified by tumor stage and patient characteristics to provide valuable insights into the prognosis and management of penile cancer.\nAccording to the basic characteristic of dead patients, people diagnosed with penile cancer between 15 and 54 years had obviously higher SMRs than other age groups. As the number of deaths among 15-54 years old were not obviously higher than other age groups, this result was due to the lower excess risk of the general population. For the general population among 15-54 years, most of them have better physical conditions and a lower rate of all kinds of deaths when compared with the population with elderly age. We also found that the mortality risk continued to decrease with the year of diagnosis from 2.65 in 2004-2008 to 2.01 in 2014-2018 in localized disease. This may be the result of improved cancer treatment and disease management after diagnosis for PeCa over time. However, this result was reversed in regional and distant metastatic disease. White people accounted for a higher proportion of all deaths, but black people had a higher risk of death in localized and regional diseases. Previous studies reported that the mortality rate of blacks was higher than that of whites among American penile cancer patients.6 This may be due to earlier exposure of blacks to HPV through early sexual contact, leading to an earlier diagnosis of penile cancer.6,7\nFor localized and regional diseases, deaths from SMR accounted for a certain percentage of total deaths. A study identified 3641 PeCa patients in the Swedish Family-Cancer Database from 1958 to 2015. They found that SMTs occurred in 16.8% of primary PeCa patients and 45.9% of these patients dead for SMTs rather than PeCa.8 The main SMTs included prostate cancer, colorectum cancer, lung cancer, bladder cancer, skin squamous cell cancer and stomach cancer. These results were similar to ours. This study also reported that SMTs was associated with HPV and smoking.8 Another study with 1634 men with PeCa found a 2-3-fold increased risk of a second human papillomavirus (HPV) associated cancer of the oral cavity, oropharynx and anal canal among PeCa patients.9 For the occurrence of PeCa, there were some important risk factors of PeCa including lack of circumcision, smegma retention, chronic balanitis, lichen sclerosis, obesity, smoking, HPV infection and immune-compromised states.10,11 It may be the consequences of treatment for penile cancer and the more widely spread of HPV.12-14\nFor the non-tumor deaths, it accounted for a large proportion among all deaths in localized and regional stages. A previous study reported that 65.7% of deaths were due to non-cancer causes among PeCa patients without SMTs and only 28.4% were due to penile cancer.8 Cardiovascular disease is still a major cause of death in PeCa patients. Statistics in recent years from a study had shown that more than 20% of PeCa patients died from cardiovascular disease.15 It was also reported that the risk of cancer patients dying from cardiovascular diseases was inversely related to the age at diagnosis.15 For this kind of cancer with a good prognosis but still a high risk of cardiovascular death, patients may benefit from clinical intervention by a cardiologist at the time of diagnosis. There were some relationships between diabetes mellitus and PeCa. Previous studies reported that diabetes was associated with phimosis, which could increase the risk of penile cancer.16-18 Men with diabetes, mainly type 2 diabetes, were associated with an increased incidence of penile squamous cell carcinoma compared with men without diabetes.19 Localized and advanced PeCa and the associated treatments have profound physical and psychological effects on the survivors’ quality of life. Due to potentially crippling surgery, patients with penile cancer increased psychological stress and therefore have an increased need for psychosocial care.20-22 However, in our study, we didn’t find a significantly increased risk of suicide and self-inflicted injury.22\nFor patients with distant metastatic PeCa, the prognosis was poor with a mortality rate of 82.6%. The main cause of death in patients was the penile cancer tumor itself. Other causes of death only account for only a small fraction of all deaths. Therefore, active and effective treatment for metastatic penile cancer is still the most important measure to prevent deaths and prolong survival durations.\nOur study is the first to provide the most comprehensive comparison and description of the risk of death from all causes in PeCa patients using tumor stage and time after diagnosis. However, there were some limitations in our study. Firstly, we used Summary Stage 2000 (1998+) to distinguish local, regional and distant diseases, but these classifications maybe not the same in different periods. Therefore, the results of this study can’t fully represent the current staging results. Secondly, some important information of patients in this database was available, such as the basic illness, detailed treatments and socioeconomic status, etc. The information might have certain impacts on the survival and prognosis of patients. Our results might be influenced by the missing information. Thirdly, the sample size selection was not estimated by power calculations. It was also a limitation of our study.","Our study provided a detailed analysis of the causes of death for patients with locally, regionally, and distant metastatic PeCa after diagnosis. This information could be useful for disease prevention and health care during patients’ survivorship.","Click here for additional data file.\nSupplemental Material for Second Malignant Tumors and Non-Tumor Causes of Death for Patients With Penile Cancer During Their Survivorship by Pan Song, Xiaotian Wu, Luchen Yang, Kai Ma, Zhenghuan Liu, Jing Zhou, Junhao Chen, Qing Zhu, and Qiang Dong in Cancer Control"],"string":"[\n \"Penile cancer (PeCa) is a rare malignancy worldwide, accounting for less than 1% of all malignancies in men.1 The overall incidence rate is approximately .69 cases per 100 000 persons in the US.2 It was estimated that the number of new PeCa patients was 34 475 and deaths cases was 15 138 each year worldwide.3 The survival rates still vary greatly with the stages of patients’ disease. It was reported that about 40% of new cases were diagnosed with localized PeCa, with a 5-year overall survival rate of approximately 90%.4 The survival rates decrease dramatically once patients are in the regional or distant metastasis stage. The 5-year overall survival rate is approximately 80% with unilateral inguinal lymph node involvement, 10%-20% with bilateral or pelvic lymph node involvement, and <10% with extranodal extension.4\\nSince the 1990s, substantial progress has been made in the prevention, diagnosis and treatment of PeCa. The prognosis has been greatly improved in PeCa patients, especially those who have not progressed to the metastatic stage. With the prolonged survival time, the proportion of other causes of death such as second malignant tumors (SMTs) and non-tumor diseases has been increasing among all deaths of PeCa patients. Since few previous studies have focused exclusively on other causes of death of PeCa patients, an analysis of causes of death among PeCa patients is warranted. Our study aimed to evaluate the main causes of death including SMTs and non-cancer causes for PeCa patients, and calculate the standardized mortality ratio (SMR) for each reason of death compared with the normal population.\",\n \"[SUBTITLE] Data Sources [SUBSECTION] The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\\nThe data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\\n[SUBTITLE] Study Population and Study Variables [SUBSECTION] Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\\nPatients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\\n[SUBTITLE] Outcome Assessments [SUBSECTION] The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\\nThe causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\\n[SUBTITLE] Statistics Analyses [SUBSECTION] For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\\nFor the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\",\n \"The data of this study was derived from the Surveillance, Epidemiology, and End Results (SEER) program 18 registries, which is an authoritative source of information on cancer incidence and survival in the United States. SEER currently collects and publishes cancer incidence and survival data from population-based cancer registries covering approximately 27.8% of the U.S. population, and it collects data concerning patient demographics, tumor morphology, stage at diagnosis, primary tumor site, the first course of treatment, and follow-up for vital status. Meanwhile, the data for the general population comes from the Centers for Disease Control and Prevention (CDC) of the United States. All data in this study has been removed identifiable information and can be available freely online. Thus, this study was regarded as exempt research by the institutional review.\",\n \"Patients with PeCa as the first primary malignancy and clear stage information according to the definition of Summary Stage 2000 (1998+) between 2004 and 2018 were included. The patients who had unclear follow-up time, living status at the end of follow-up and detailed causes of death were excluded.\\nPatients were divided into 3 groups by their stages (localized, regional and distant metastasis). The following variables of all patients and the dead patients were analyzed, including age (15-54, 55-64, 65-74, 75-84, and >85 years), year of diagnosis (2004-2008, 2009-2013, and 2014-2018), race (white, black. American Indian/Alaska native and Asian or Pacific Islander), grade (well-differentiated, moderately differentiated, poorly differentiated, and undifferentiated), main pathological type, Radiotherapy (beam radiation, radioactive implants/brachytherapy and none/unknown) and Chemotherapy (yes and no/unknown).\",\n \"The causes of death after the diagnosis of PeCa included PeCa deaths, SMTs, and non-cancer deaths were regarded as the main outcomes. The standardized mortality ratio (SMR) with the 95% confidence interval (95% CI) of each cause was also calculated to evaluate the mortality risks compared with the general population. The last follow-up time was December 31, 2018, which was the latest data updated in the SEER database.\",\n \"For the baseline characteristics of 3 groups of patients with different PeCa stages, we analyzed the number of all included patients and observed deaths stratified by the above variables. The SMRs for each variable were calculated by comparing the observed number of deaths to the expected number. The expected number of deaths was based on the total number of patient-year and the excess risk of the general population. The causes of death and related SMRs and 95% CI were also analyzed in localized, regional, and distant metastasis PeCa patients. Statistical significance for SMR is a two-sided test of P < .05. All these analyses were performed using SEER*Stat version 8.4.0.1. Pie charts of the proportion of different causes of death were drawn with Microsoft Excel 2016.\",\n \"A total of 3597 PeCa patients were included, 1588 (44%) of them died during the follow-up. The majority of patients was in the localized stage (n = 2,155, 60%) and regional stage (n = 1310, 36%), whereas only 4% (n = 132) had distant disease. For patients with localized PeCa, 800 (37.1%) of them died with a SMR of 2.12 (1.97-2.27). 679 (51.8%) regional PeCa patients [SMR: 4.61 (4.27-4.97)] and 109 (82.6%) patients with distant metastatic PeCa [SMR: 27.68 (22.73-33.39)] died during the follow-up.\\nBaseline characteristics of included patients and the dead cases were presented in Table 1. The trend of new diagnosed PeCa patients from 2000 to 2018 were revealed in Figure 1.Table 1.Baseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.CharacteristicLocalizedRegionalDistantPatients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Patients, nDeaths, nSMR (95% CI)Total21558002.12a (1.97-2.27)13106794.61a (4.27-4.97)13210927.68a (22.73-33.39)Age 15-54 years470804.41a (3.5-5.49)32912116.05a (13.32-19.18)3630106.43a (71.81-151.93) 55-64 years4991282.86a (2.39-3.4)3371537.48a (6.34-8.77)322748.63a (32.05-70.75) 65-74 years5652102.21a (1.92-2.53)3301834.67a (4.01-5.39)332635.53a (23.21-52.06) 75-84 years4102221.55a (1.35-1.76)2171432.61a (2.2-3.08)262110.51a (6.51-16.07) 85+ years2111602.10a (1.79-2.45)97793.10a (2.45-3.86)5513.45a (4.37-31.39)Year of diagnosis 2004-20087301432.65a (2.23-3.12)3532343.11a (2.73-3.54)302717.20a (11.34-25.03) 2009-20137442842.05a (1.82-2.3)4422585.32a (4.69-6.01)484232.68a (23.55-44.17) 2014-20186813732.01a (1.81-2.23)5151877.88a (6.79-9.1)544036.91a (26.37-50.27)Race White18066782.08a (1.92-2.24)11205774.35a (4-4.72)1058326.90a (21.43-33.35) Black222842.48a(1.98-3.07)116727.28a (5.7-9.17)161520.92a (11.71-34.51) American Indian/Alaska native2183.79a (1.64-7.47)1137.25a (1.49-21.18)33586.91a (121.03-1715.19) Asian or Pacific islander106301.95a (1.31-2.78)63276.22a (4.1-9.05)8861.02a (26.35-120.24)Grade Well differentiated6312181.78a (1.55-2.03)188863.01a (2.41-3.71)7620.27a (7.44-44.12) Moderately differentiated7063042.55a (2.27-2.85)5743175.07a (4.52-5.65)383219.92a (13.62-28.11) Poorly differentiated2571302.93a (2.44-3.47)3171925.02a (4.34-5.79)444135.67a (25.6-48.39) Undifferentiated92.7 (.08-2.52)1175.63a (2.27-11.61)329.89a (1.2-35.72)Pathological type 8070/3: Squamous cell carcinoma, NOS14255602.22a (2.04-2.41)8294474.49a (4.08-4.93)846830.34a (23.56-38.46) 8071/3: Squamous cell carcinoma, keratinizing, NOS3471262.26a (1.88-2.69)3321655.54a (4.73-6.46)191414.96a (8.18-25.1) 8051/3: Verrucous carcinoma, NOS169491.50a (1.11-1.98)39101.54 (.74-2.84)2298.29a (11.9-355.06) 8083/3: Basaloid squamous cell carcinoma41122.96a (1.53-5.18)34165.50a(3.14-8.93)6523.04a (7.48-53.77)Radiotherapy Beam radiation68301.90a (1.28-2.72)1981288.26a(6.89-9.83)372945.93a(30.76-65.96) Radioactive implants/brachytherapy214.79 (.12-26.7)3112.38 (.31-68.98)000 None/unknown20857692.13a (1.98-2.28)11095504.17a(3.83-4.54)958024.19a(19.18-30.11)Chemotherapy Yes59222.69a (1.69-4.08)29816711.38a(9.72-13.24)745831.87a(24.2-41.21) No/unknown20967782.10a (1.96-2.26)10125123.86a(3.53-4.21)585124.07a(17.92-31.65)SMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.aStatistical significance with P < .05.Figure 1.The number of newly diagnosed of penile cancer from 2004 to 2018.\\nBaseline Characteristics of All Patients and Dead Cases During Follow-Up With Localized, Regional, and Distant Penile Cancer.\\nSMR: standardized mortality ratio; 95% CI: 95% Confidence interval; NOS: Not otherwise specified.\\naStatistical significance with P < .05.\\nThe number of newly diagnosed of penile cancer from 2004 to 2018.\\n[SUBTITLE] Causes of Death for Patients With Localized Penile Cancer [SUBSECTION] Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nMain causes of death for patients with localized penile cancer.\\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\\naStatistical significance with P < .05.\\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nAmong the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nMain causes of death for patients with localized penile cancer.\\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\\naStatistical significance with P < .05.\\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\n[SUBTITLE] Causes of Death for Patients With Regional Penile Cancer [SUBSECTION] For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\\nMain causes of death for patients with regional penile cancer.\\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\\naStatistical significance with P < .05.\\nFor patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\\nMain causes of death for patients with regional penile cancer.\\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\\naStatistical significance with P < .05.\\n[SUBTITLE] Causes of Death for Patients With Distant Metastatic Penile Cancer [SUBSECTION] For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\\nMain Causes of Death for Patients With Distant Penile Cancer.\\nSMR: standardized mortality ratio; CI: confidence interval.\\naStatistical significance with P < .05.\\nFor the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\\nMain Causes of Death for Patients With Distant Penile Cancer.\\nSMR: standardized mortality ratio; CI: confidence interval.\\naStatistical significance with P < .05.\",\n \"Among the 800 cases of death, 199 (24.9%) died for PeCa, 144 (18.0%) died for other tumor causes, and the rest 457 (57.1%) deaths were due to non-tumor causes. 75.9% of all PeCa deaths were within 3 years after diagnosis. The mortality rate of the general population who died for PeCa was rather low, which led to the extremely high SMR from the ratio of observed and expected. The majority of SMTs deaths (n = 106, 73.6%) occurred with 5 years after the diagnosis of PeCa. The main SMTs were lung and bronchus cancer [n = 40, SMR: 1.71 (1.22-2.33)], all cancers of digestive system [n = 25, SMR: 1.12 (.72-1.65)]. The main non-cancer causes of death were diseases of heart [n = 172, SMR: 1.66 (1.42-1.93)], chronic obstructive pulmonary disease (COPD) and allied cond [n = 38, SMR: 1.63 (1.15-2.24)], cerebrovascular diseases [n = 33, 22% of all deaths, SMR: 1.71 (1.17-2.4)] and diabetes mellitus [n = 26, SMR: 2.30 (1.5-3.37)]. The mortality risks of lung and bronchus cancer, skin tumor, diseases of heart, COPD and allied cond, cerebrovascular diseases, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis, other infectious and parasitic diseases, hypertension, chronic liver disease and cirrhosis and atherosclerosis were significantly increased when compared with the general population. All these results are shown in Table 2. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes of patients after diagnosis is shown in Figure 2A and the percentage composition of the main causes of death is presented in Figure 3A.Table 2.Main causes of death for patients with localized penile cancer.Causes of deathTotal<1 year<1-5 year5-10 years>10 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death800377.822.12a (1.97-2.27)16758.342.86a (2.44-3.33)422183.122.30a (2.09-2.54)167107.681.55a (1.32-1.8)4428.681.53a (1.11-2.06)All malignant cancers34386.53.97a (3.56-4.41)8013.455.95a (4.72-7.4)20442.354.82a (4.18-5.53)4824.481.96a (1.45-2.6)116.221.77 (.88-3.16) Male genital system21310.7219.86a (17.29-22.72)581.7133.97a (25.79-43.91)1295.224.82a (20.72-29.49)252.998.37a (5.42-12.36)1.831.2 (.03-6.7) Penis199.082495.58a (2160.87-2867.44)53.014477.48a (3353.94-5856.66)125.043293.53a (2741.5-3924.09)21.02898.28a (556.05-1373.11)0.010 (0-561.23) Prostate910.59.85 (.39-1.61)21.691.18 (.14-4.28)25.14.39 (.05-1.41)42.951.36 (.37-3.47)1.821.22 (.03-6.78) Urinary system86.291.27 (.55-2.51)3.963.14 (.65-9.16)43.051.31 (.36-3.36)11.81.55 (.01-3.08)0.470 (0-7.81) Urinary bladder43.81.05 (.29-2.7)2.583.47 (.42-12.54)11.83.55 (.01-3.04)11.1.91 (.02-5.09)0.290 (0-12.7) Kidney and renal Pelvis12.33.43 (.01-2.39)1.362.81 (.07-15.63)01.140 (0-3.24)0.670 (0-5.54)0.170 (0-21.87) Other urinary organs3.0933.47a (6.9-97.82)0.010 (0-278.54)3.0469.99a (14.43-204.54)0.030 (0-140.81)0.010 (0-503.78) Digestive system2522.341.12 (.72-1.65)33.41.88 (.18-2.57)1810.891.65 (.98-2.61)36.4.47 (.1-1.37)11.63.61 (.02-3.42) Liver and intrahepatic bile duct83.712.16 (.93-4.25)0.530 (0-6.92)71.783.93a (1.58-8.09)11.11.9 (.02-5.03)0.280 (0-13.01) Pancreas75.471.28 (.51-2.63)1.821.22 (.03-6.82)52.651.89 (.61-4.4)11.59.63 (.02-3.5)0.420 (0-8.88) Esophagus42.931.37 (.37-3.5)0.450 (0-8.21)31.442.08 (.43-6.09)0.840 (0-4.41)1.214.84 (.12-26.98) Colon and rectum37.41.4 (.08-1.18)21.181.7 (.21-6.13)03.650 (0-1.01)12.06.48 (.01-2.7)0.520 (0-7.13) Stomach21.831.1 (.13-3.96)0.290 (0-12.81)20.92.23 (.27-8.06)0.510 (0-7.2)0.130 (0-28.41) Intrahepatic bile duct2.762.63 (.32-9.49)0.110 (0-34.14)1.362.77 (.07-15.41)1.234.36 (.11-24.32)0.060 (0-59.14) Respiratory system4324.311.77a (1.28-2.38)73.871.81 (.73-3.73)2512.072.07a (1.34-3.06)86.761.18 (.51-2.33)31.621.85 (.38-5.41) Lung and bronchus4023.391.71a (1.22-2.33)73.721.88 (.76-3.87)2411.622.07a (1.32-3.07)66.5.92 (.34-2.01)3z1.93 (.4-5.64) Skin excluding basal and squamous112.284.82a (2.41-8.63)1.352.88 (.07-16.03)61.125.38a (1.97-11.7)3.664.57 (.94-13.37)1.166.17 (.16-34.36) Brain and other nervous system21.761.14 (.14-4.11)0.260 (0-13.99)2.862.34 (.28-8.44)0.510 (0-7.23)0.130 (0-28.57) Lymphoma33.5.86 (.18-2.5)0.540 (0-6.79)11.71.58 (.01-3.25)2.992.01 (.24-7.27)0.250 (0-14.66) Myeloma21.931.04 (.13-3.75)0.290 (0-12.63)1.931.07 (.03-5.97)1.551.81 (.05-10.08)0.150 (0-24.97) Leukemia33.89.77 (.16-2.25)00.60 (0-6.15)21.91.05 (.13-3.8)11.11.9 (.02-5.02)0.280 (0-13.08) Miscellaneous malignant cancer286.794.12a (2.74-5.96)71.056.64a (2.67-13.67)133.323.91a (2.08-6.69)41.922.08 (.57-5.32)4.498.13#(2.21-20.81)Non-cancer causes Diseases of heart172103.441.66a (1.42-1.93)2916.311.78a (1.19-2.55)9150.311.81a (1.46-2.22)4029.081.38 (.98-1.87)127.741.55 (.8-2.71) COPD and allied cond3823.291.63a (1.15-2.24)83.592.23 (.96-4.4)2011.321.77a (1.08-2.73)96.651.35 (.62-2.57)11.74.57 (.01-3.2) Cerebrovascular diseases3319.341.71#(1.17-2.4)73.042.31 (.93-4.75)109.331.07 (.51-1.97)95.461.65 (.75-3.13)71.514.63#(1.86-9.54) Diabetes mellitus2611.322.30a (1.5-3.37)61.723.49#(1.28-7.6)75.471.28 (.51-2.64)113.253.38a (1.69-6.05)2.882.28 (.28-8.24) Nephritis, nephrotic syndrome and nephrosis168.251.94a (1.11-3.15)31.282.34 (.48-6.83)1142.75a (1.37-4.92)12.33.43 (.01-2.39)1.631.58 (.04-8.79) Pneumonia and influenza149.331.5 (.82-2.52)51.53.34a (1.08-7.79)64.541.32 (.49-2.88)32.61.16 (.24-3.38)00.70 (0-5.3) Other infectious and parasitic diseases including HIV133.134.16a (2.21-7.11)5.4910.17a (3.3-23.74)31.551.93 (.4-5.64)5.885.71a (1.85-13.33)0.210 (0-17.77) Accidents and adverse effects1112.11.91 (.45-1.63)21.821.1 (.13-3.96)45.84.69 (.19-1.75)33.52.85 (.18-2.49)2.932.16 (.26-7.79) Hypertension without heart disease104.222.37#(1.14-4.36)2.623.22 (.39-11.62)11.99.5 (.01-2.8)51.244.04a (1.31-9.42)2.365.52 (.67-19.96) Chronic liver disease and cirrhosis93.82.37a (1.08-4.49)2.573.53 (.43-12.76)41.852.16 (.59-5.53)21.111.8 (.22-6.49)1.273.65 (.09-20.36)AbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.aStatistical significance with P < .05.Figure 2.Mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.Figure 3.The percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nMain causes of death for patients with localized penile cancer.\\nAbbreviationsSMR: standardized mortality ratio; CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease.\\naStatistical significance with P < .05.\\nMortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients diagnosed with localized, regional and distant metastasis penile cancer. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\\nThe percentages of main causes of death in penile cancer patients. (A) Patients with localized penile cancer; (B) patients with regional penile cancer; (C) patients with distant metastasis penile cancer.\",\n \"For patients with regional PeCa, 348 (51.3%) of all 679 deaths were due to PeCa itself. There were 118 SMTs deaths and 213 non-cancer deaths, accounting for 17.4% and 31.4% of all deaths. For the PeCa deaths, 89.1% of them occurred within 3 years after diagnosis. With the low PeCa mortality rate of the general population, the SMR of PeCa death reached a terribly high level. The main SMTs were lung and bronchus cancer [n = 23, SMR: 2.41 (1.53-3.62)]. The mortality risks of lung and bronchus cancer, and skin tumor [SMR: 6.41 (2.35-13.95)] were significantly increased. The most common non-cancer cause of death was diseases of heart [n = 71, SMR: 1.77 (1.38-2.23)], COPD and allied cond [n = 17, SMR: 1.85 (1.08-2.95)], diabetes mellitus [n = 16, SMR: 3.62 (2.07-5.88)] and cerebrovascular diseases [n = 13, SMR: 1.77 (.94-3.03)]. The mortality risks of diseases of heart, COPD and allied cond, diabetes mellitus, nephritis, nephrotic syndrome and nephrosis [n = 9, SMR: 2.88 (1.32-5.47)] and septicemia [n = 8, SMR: 3.83 (1.65-7.54)] were significantly increased when compared with the general population. All these results are shown in Table 3. The mortality rate of patients after diagnosis and is shown in Figure 2B and the percentages of the main causes of death is revealed in Figure 3B.Table 3.Main causes of death for patients with regional penile cancer.Causes of deathTotal<1 year1-5 years>5 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death679147.44.61a (4.27-4.97)27527.559.98a (8.84-11.23)31067.934.56a (4.07-5.1)9451.921.81a (1.46-2.22)All malignant cancers46634.6513.45a (12.26-14.73)2136.5632.48a (28.27-37.15)22516.4213.71a (11.97-15.62)2811.672.40a (1.59-3.47) Male genital system3624.0788.89a (79.97-98.54)179.79225.26a (193.47-260.79)1681.8889.54a (76.51-104.15)151.410.70a (5.99-17.65) Penis348.0311 020.00a (9892.46-12,240.84)173.0129 605.21a (25 357.89-34 360.42)162.0111 182.19a (9526.53-13042.87)13.011155.76a (615.39-1976.38) Prostate84.021.99 (.86-3.92)2.782.55 (.31-9.21)41.852.16 (.59-5.53)21.381.45 (.18-5.22) Testis2.01149.35a (18.09-539.5)10370.55a (9.38-2064.6)1.01152.52a (3.86-849.81)000 (0-891.78) Urinary system62.512.39 (.88-5.21)1.462.18 (.06-12.16)41.153.47 (.94-8.87)1.891.12 (.03-6.23) Kidney and renal Pelvis3.953.17 (.65-9.27)0.180 (0-20.85)2.454.46 (.54-16.13)1.323.12 (.08-17.36) Urinary bladder21.51.34 (.16-4.83)1.273.71 (.09-20.65)1.681.48 (.04-8.23)0.550 (0-6.72) Digestive system78.97.78 (.31-1.61)11.69.59 (.01-3.3)44.27.94 (.26-2.4)23.01.66 (.08-2.4) Colon and rectum32.951.02 (.21-2.97)0.570 (0-6.44)21.411.41 (.17-5.11)1.971.03 (.03-5.77) Esophagus21.211.65 (.2-5.95)0.230 (0-16.24)1.581.72 (.04-9.57)1.412.47 (.06-13.75) Respiratory system239.92.32a (1.47-3.49)41.912.1 (.57-5.37)134.82.71a (1.44-4.63)63.191.88 (.69-4.09) Lung and bronchus239.532.41a (1.53-3.62)41.842.18 (.59-5.58)134.622.81a (1.5-4.81)63.071.95 (.72-4.25) Skin excluding basal and squamous6.946.41a (2.35-13.95)2.1711.79a (1.43-42.6)4.439.23a (2.51-23.62)0.330 (0-11.07) Lymphoma21.411.42 (.17-5.14)0.260 (0-14.17)2.653.05 (.37-11.03)0.490 (0-7.53) Leukemia21.561.28 (.16-4.63)0.290 (0-12.84)2.722.76 (.33-9.97)0.550 (0-6.74) Miscellaneous malignant cancer512.7218.75a (13.96-24.66)25.5148.95a (31.68-72.26)241.2818.76a (12.02-27.92)2.932.15 (.26-7.77)Non-cancer causes Diseases of heart7140.111.77a (1.38-2.23)207.572.64#(1.61-4.08)3218.451.73a (1.19-2.45)1914.091.35 (.81-2.11) COPD and allied cond179.211.85a (1.08-2.95)41.692.37 (.65-6.08)84.241.89 (.82-3.72)53.291.52 (.49-3.55) Diabetes mellitus164.423.62a (2.07-5.88)5.836.00a (1.95-14.01)62.072.89a (1.06-6.3)51.513.31a (1.08-7.73) Cerebrovascular diseases137.341.77 (.94-3.03)31.392.16 (.44-6.3)63.341.79 (.66-3.91)42.611.53 (.42-3.92) Nephritis, nephrotic syndrome and nephrosis93.132.88a (1.32-5.47)3.595.07a (1.05-14.81)51.443.48a (1.13-8.12)11.1.91 (.02-5.08) Septicemia82.093.83a (1.65-7.54)50.412.66a (4.11-29.53)1.961.04 (.03-5.77)2.732.74 (.33-9.89) Alzheimers84.791.67 (.72-3.29)3.843.56 (.74-10.42)22.02.99 (.12-3.57)31.931.56 (.32-4.55) Chronic liver disease and cirrhosis41.622.47 (.67-6.32)3.319.67a (1.99-28.25)0.790 (0-4.64)1.521.94 (.05-10.8) Accidents and adverse effects64.841.24 (.45-2.7)0.910 (0-4.04)32.261.33 (.27-3.88)31.671.8 (.37-5.25) Suicide and self-inflicted injury41.382.9 (.79-7.42)1.273.71 (.09-20.68)1.671.49 (.04-8.28)2.444.57 (.55-16.51)CI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.aStatistical significance with P < .05.\\nMain causes of death for patients with regional penile cancer.\\nCI: confidence interval; COPD: Chronic Obstructive Pulmonary Disease; SMR: standardized mortality ratio.\\naStatistical significance with P < .05.\",\n \"For the included 132 patients with distant metastasis PeCa, 109 of them died during the follow-up with a significantly increased SMR of 27.68 (22.73-33.39). 100 deaths (91.7%) were due to malignant cancers including 76 cases (69.7%) dead of PeCa. 67.1% (n = 51) of all PeCa deaths occurred within 1 year and 98.7% (n = 75) dead within 3 years after diagnosis. There were 24 patients (22%) who died for SMTs, mainly including cancers from 4 lung and bronchus, 2 prostate, and 2 skin, 2 lymphoma and 3 other urinary organs. 19 (79.2%) of SMTs deaths occurred within 1 year of the diagnosis of PeCa. As for the non-cancer deaths, only 9 cases of deaths were observed, including 5 deaths from diseases of heart, 1 septicemia death, 1 other infection and parasitic diseases deaths, and 2 other causes of death. These results are presented in Table 4 and Figure 3C. The mortality rate of all causes of death, PeCa, SMTs and non-tumor causes for patients after diagnosis is shown in Figure 2C.Table 4.Main Causes of Death for Patients With Distant Penile Cancer.Causes of deathTotal<1 year1-3 years>3 yearsObserved, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)Observed, nExpected, nSMR (95% CI)All causes of death1093.9427.68a (22.73-33.39)741.8540.10a (31.49-50.34)281.0925.61a (17.02-37.01)717.00a (2.82-14.43)All malignant cancers100.98102.30a (83.23-124.42)70.45154.76a (120.65-195.54)27.2992.21a (60.76-134.16)3.2312.91a (2.66-37.72) Respiratory system4.2814.38a (3.92-36.81)3.1322.93a (4.73-67.01)0.090 (0-42.56)1.0616.48 (.42-91.8) Lung and bronchus4.2714.98a (4.08-38.34)3.1323.85a (4.92-69.69)0.080 (0-44.28)1.0617.24 (.44-96.06) Soft tissue including heart2.01331.85a (40.19-1198.77)10351.01a (8.89-1955.69)10554.76a (14.05-3090.9)000 (0-2682.36) Skin excluding basal and squamous2.0285.91a (10.4-310.32)1.0187.34a (2.21-486.61)1.01146.53a (3.71-816.42)0.010 (0-736.74) Male genital system80.12650.98a (516.18-810.2)55.051065.43a (802.63-1386.8)24.03713.33a (457.04-1061.38)1.0426.58 (.67-148.08) Prostate2.1216.48a (2-59.52)2.0539.26a (4.75-141.82)0.030 (0-111.09)0.040 (0-99.09) Penis76084 751.04a (66 774.18-106078.56)510126 908.70a (94 491.86-166861.5)24092 697.90a (59 393.29-137,927.13)104237.74a (107.29-23,611.17) Urinary system3.0646.59a (9.61-136.15)2.0364.45a (7.81-232.82)0.020 (0-199.1)1.0167.40a (1.71-375.54) Miscellaneous malignant cancer7.0793.72a (37.68-193.1)7.04198.90a (79.97-409.8)0.020 (0-166.27)0.020 (0-213.09)Non-cancer causes Septicemia1.0616.57 (.42-92.3)1.0337.63 (.95-209.66)0.020 (0-218.26)0.020 (0-218.46) Other infectious and parasitic diseases including HIV1.0426.85 (.68-149.58)1.0257.29a (1.45-319.22)0.010 (0-340.04)0.010 (0-412.29) Diseases of heart51.054.75a (1.54-11.09)1.492.02 (.05-11.26)1.283.52 (.09-19.64)3.2710.95a (2.26-32) Other cause of death2.573.53 (.43-12.75)1.273.69 (.09-20.57)0.150 (0-25.01)1.156.75 (.17-37.6)SMR: standardized mortality ratio; CI: confidence interval.aStatistical significance with P < .05.\\nMain Causes of Death for Patients With Distant Penile Cancer.\\nSMR: standardized mortality ratio; CI: confidence interval.\\naStatistical significance with P < .05.\",\n \"With the high and increasing survival rate of PeCa, deaths from other causes account for a very high proportion of total deaths and tend to increase.5 Therefore, other causes of death including SMTs and non-tumor diseases should be taken seriously. Besides, whether these causes have relationships with the diagnosis of PeCa is unclear. In current literature, studies on the SMTs for the primary PeCa were quite rare. Our study, as far as we can know, is the first study focusing on the SMTs and non-cancer causes of death for primary PeCa patients. We also relate these causes to the time period after diagnosis of penile cancer and provide results stratified by tumor stage and patient characteristics to provide valuable insights into the prognosis and management of penile cancer.\\nAccording to the basic characteristic of dead patients, people diagnosed with penile cancer between 15 and 54 years had obviously higher SMRs than other age groups. As the number of deaths among 15-54 years old were not obviously higher than other age groups, this result was due to the lower excess risk of the general population. For the general population among 15-54 years, most of them have better physical conditions and a lower rate of all kinds of deaths when compared with the population with elderly age. We also found that the mortality risk continued to decrease with the year of diagnosis from 2.65 in 2004-2008 to 2.01 in 2014-2018 in localized disease. This may be the result of improved cancer treatment and disease management after diagnosis for PeCa over time. However, this result was reversed in regional and distant metastatic disease. White people accounted for a higher proportion of all deaths, but black people had a higher risk of death in localized and regional diseases. Previous studies reported that the mortality rate of blacks was higher than that of whites among American penile cancer patients.6 This may be due to earlier exposure of blacks to HPV through early sexual contact, leading to an earlier diagnosis of penile cancer.6,7\\nFor localized and regional diseases, deaths from SMR accounted for a certain percentage of total deaths. A study identified 3641 PeCa patients in the Swedish Family-Cancer Database from 1958 to 2015. They found that SMTs occurred in 16.8% of primary PeCa patients and 45.9% of these patients dead for SMTs rather than PeCa.8 The main SMTs included prostate cancer, colorectum cancer, lung cancer, bladder cancer, skin squamous cell cancer and stomach cancer. These results were similar to ours. This study also reported that SMTs was associated with HPV and smoking.8 Another study with 1634 men with PeCa found a 2-3-fold increased risk of a second human papillomavirus (HPV) associated cancer of the oral cavity, oropharynx and anal canal among PeCa patients.9 For the occurrence of PeCa, there were some important risk factors of PeCa including lack of circumcision, smegma retention, chronic balanitis, lichen sclerosis, obesity, smoking, HPV infection and immune-compromised states.10,11 It may be the consequences of treatment for penile cancer and the more widely spread of HPV.12-14\\nFor the non-tumor deaths, it accounted for a large proportion among all deaths in localized and regional stages. A previous study reported that 65.7% of deaths were due to non-cancer causes among PeCa patients without SMTs and only 28.4% were due to penile cancer.8 Cardiovascular disease is still a major cause of death in PeCa patients. Statistics in recent years from a study had shown that more than 20% of PeCa patients died from cardiovascular disease.15 It was also reported that the risk of cancer patients dying from cardiovascular diseases was inversely related to the age at diagnosis.15 For this kind of cancer with a good prognosis but still a high risk of cardiovascular death, patients may benefit from clinical intervention by a cardiologist at the time of diagnosis. There were some relationships between diabetes mellitus and PeCa. Previous studies reported that diabetes was associated with phimosis, which could increase the risk of penile cancer.16-18 Men with diabetes, mainly type 2 diabetes, were associated with an increased incidence of penile squamous cell carcinoma compared with men without diabetes.19 Localized and advanced PeCa and the associated treatments have profound physical and psychological effects on the survivors’ quality of life. Due to potentially crippling surgery, patients with penile cancer increased psychological stress and therefore have an increased need for psychosocial care.20-22 However, in our study, we didn’t find a significantly increased risk of suicide and self-inflicted injury.22\\nFor patients with distant metastatic PeCa, the prognosis was poor with a mortality rate of 82.6%. The main cause of death in patients was the penile cancer tumor itself. Other causes of death only account for only a small fraction of all deaths. Therefore, active and effective treatment for metastatic penile cancer is still the most important measure to prevent deaths and prolong survival durations.\\nOur study is the first to provide the most comprehensive comparison and description of the risk of death from all causes in PeCa patients using tumor stage and time after diagnosis. However, there were some limitations in our study. Firstly, we used Summary Stage 2000 (1998+) to distinguish local, regional and distant diseases, but these classifications maybe not the same in different periods. Therefore, the results of this study can’t fully represent the current staging results. Secondly, some important information of patients in this database was available, such as the basic illness, detailed treatments and socioeconomic status, etc. The information might have certain impacts on the survival and prognosis of patients. Our results might be influenced by the missing information. Thirdly, the sample size selection was not estimated by power calculations. It was also a limitation of our study.\",\n \"Our study provided a detailed analysis of the causes of death for patients with locally, regionally, and distant metastatic PeCa after diagnosis. This information could be useful for disease prevention and health care during patients’ survivorship.\",\n \"Click here for additional data file.\\nSupplemental Material for Second Malignant Tumors and Non-Tumor Causes of Death for Patients With Penile Cancer During Their Survivorship by Pan Song, Xiaotian Wu, Luchen Yang, Kai Ma, Zhenghuan Liu, Jing Zhou, Junhao Chen, Qing Zhu, and Qiang Dong in Cancer Control\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["intro","methods",null,null,null,null,"results",null,null,null,"discussion","conclusion","supplementary-material"],"string":"[\n \"intro\",\n \"methods\",\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\",\n \"conclusion\",\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["penile cancer","causes of death","second malignant tumors","non-tumor diseases","standardized mortality ratio","SEER"],"string":"[\n \"penile cancer\",\n \"causes of death\",\n \"second malignant tumors\",\n \"non-tumor diseases\",\n \"standardized mortality ratio\",\n \"SEER\"\n]"}}},{"rowIdx":379,"cells":{"title":{"kind":"string","value":"Qualitative Analysis of the Roles of Physicians and Nurses in Providing Decision Support to Patients With Relapsed or Refractory Leukemia and Lymphoma."},"pmid":{"kind":"string","value":"36268680"},"background_abstract":{"kind":"string","value":"This study examined the roles of hematologists and other professionals in providing decision support to patients with relapsed or refractory leukemia and lymphoma."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"This was a qualitative study using in-depth semi-structured interviews involving 11 hematologists in Japan."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"We identified 7 categories related to the roles of hematologists in providing direct decision support to patients: (1) preparing patients before informed consent, (2) selecting the information to convey, (3) choosing a method for conveying this information, (4) respecting the intentions of patients and their families, (5) directing decision-making and considering fairness, (6) considering the emotional aspects of patients and their families, and (7) providing support after discussing treatment options. We also identified the following 5 subcategories related to the roles of hematologists in multidisciplinary collaboration: (1) communicating with other professionals, (2) gathering information from them, (3) providing information to them, (4) managing the entire medical team, and (5) encouraging nurses to actively participate with patients throughout the decision-making process."},"results_abstract_label":{"kind":"string","value":"RESULT"},"conclusions_abstract":{"kind":"string","value":"Through content analysis, the hematologist's direct role in decision-making was extracted as preparation and consideration in situations where information about decision-making is communicated, and emotional support after the information is communicated. In addition, active participation in discussions, sharing information about the patient's situation and relevant discussions, and emotional support as the hematologist's expected roles in other professions were extracted. The results therefore suggest that a multidisciplinary team is needed to share information and provide multidimensional support to patients."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Communication","Physicians","Lymphoma","Qualitative Research","Leukemia","Decision Making"],"string":"[\n \"Humans\",\n \"Communication\",\n \"Physicians\",\n \"Lymphoma\",\n \"Qualitative Research\",\n \"Leukemia\",\n \"Decision Making\"\n]"},"pmcid":{"kind":"string","value":"9597203"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"The prognosis and illness trajectory of leukemia and lymphoma vary greatly depending\non the degree of progression and subtype of the disease. However, irrespective of\nthe illness trajectory, patients face the same issues and debates related to\ndecision-making, such as how far to go with aggressive treatment during the relapse\nor treatment-resistant period and the choice of treatment and place of\ncare.1-3 Even in relapse cases, such\npatients may have the option to receive high-dose chemotherapy treatments or\nhematopoietic cell transplantations aimed at remission and achieving long-term\nsurvival. Such treatment carries significant risks, including complications such as\nsevere infections and bleeding due to bone marrow suppression and treatment-related\ndeath, and may result in transition from invasive treatment aimed at cure to\nend-of-life (EOL). Even when the disease is refractory to treatment, anticancer\ntherapy aimed at disease control is often administered until the end of life. In\nsuch contexts, it is very difficult for patients and their families to make\ndecisions about these complex treatment strategies and the high risks involved in\nthe context of their own life values.4-6\nThe American Society of Clinical Oncology (ASCO) has published consensus guidelines\nfor clinician-patient communication, specifically recommending that clinicians\nprovide information that is congruent with their patients’ interests and intentions\nwhile jointly considering their goals for treatment.7 It is also notable that\nhematologists’ communication skills have been receiving increasing attention in\nhematology clinical practice in recent years. An ethnographic study of\nhematologists' communication patterns when communicating bad news revealed 4\ncharacteristics: (1) technical-defensive patterns, (2) an authoritative pattern, (3)\na relational-recursive pattern, and (4) a compassionate sharing pattern, and the\npaper also suggested that hematologists were not good at expressing caring and\nempathy.8 In contrast, a qualitative study examining the needs of patients\nwith hematological malignancies when being told bad news reported identifying\nemotional support from their healthcare providers and information needs about future\nsupport systems. It was also noted that healthcare-team involvement is necessary to\nmeet these patients' needs and support their decision-making.9\nTherefore, this study sought to clarify the following: (a) the way in which\nhematologists perceive the roles of doctors in providing decision-making support to\npatients suffering from relapsed and refractory leukemia as well as lymphoma, (b)\nthe expected roles of hematologists in supporting decision-making via\nmultidisciplinary collaboration, and (c) the roles hematologists expect other\nhealthcare providers to fulfill."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"The purpose of this study was to determine hematologists’ role in decision support\nfor patients with leukemia and lymphoma. Physicians at institutions with hematology\nwards and hematology outpatient clinics where chemotherapy and bone marrow\ntransplantation are performed were included. The eligibility criteria for physicians\nincluded being Board Certified Hematologists and having at least 5 years of\nexperience. Hematologists meeting these criteria were considered to have experience\nseeing patients with relapsed or refractory leukemia or lymphoma. Subjects were\nselected using opportunistic sampling, which is used in selecting those with common\ncharacteristics and the potential to provide a rich, appropriate, and diverse\nresearch agenda. Eligible physicians were asked to recommend a qualified physician\nwho met the inclusion criteria to the head of the hematology department of a\nhospital in Tokyo. The recommended physicians were informed orally and in writing\nabout the purpose and methods of the study and the voluntary participation clause.\nAlthough this is a qualitative study and it is difficult to theoretically calculate\nthe sample size, we chose a sample size of 10 to 15 participants based on previous\nstudies9-11 using the\nsame content analysis method as in this study. The acceptance rate of recommended\nphysicians was 100%.\nVerbal and written explanations of the research objectives and procedures were\nprovided to all participants, and written consent was obtained. This study was\napproved by the Institutional Review Board of University of Tokyo (Approval No.\n3443). The study was initiated at the University of Tokyo and completed at Kyoto\nUniversity.\nTo collect the data, we conducted in-depth interviews to identify the attributes and\nroles of doctors as well as the expected roles of nurses in providing\ndecision-making support to patients with relapsed or refractory leukemia and\nlymphoma. Specifically, we scheduled face-to-face semi-structured interviews in\nprivate rooms between August and December 2011. All the interviews conducted were\naudio recorded and fully transcribed by 1 researcher (a research nurse), who also\ntook detailed field notes during each interview. Physicians were first asked for\ntheir demographic details, including age, the type of facility where they worked,\nand the number of years they had worked in their area of specialization. We then\nasked them about their roles in treating patients suffering from relapsed or\nrefractory leukemia and lymphoma.\nTo analyze the data, interview data were transcribed verbatim from the audio\nrecordings. Interview data were then numbered for each participant, separated from\npersonal information, and anonymized. A content analysis was then conducted. First,\n1 researcher extracted all statements related to the 3 study topics from each\ntranscript (eg, the role of hematologists, the role of hematologists in\nmultidisciplinary collaboration, and the expectations of other professionals in the\ncontext of providing decision-making support). We then carefully conceptualized the\ntext and categorized it into content areas through content analysis using the Klaus\nKrippendorff12 method. The units, codes, and categories were decided\nthrough consensus. Next, another experienced qualitative researcher read the\noriginal interviews and classified them into categories and subcategories. The coder\nread the original interview text and independently determined the subcategory into\nwhich the content fell. The rate of agreement between the researchers and coders was\n91%. Finally, the coder and the researcher discussed the units they disagreed on\nuntil they reached a consensus."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Participant Characteristics [SUBSECTION] We conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).\nWe conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).\n[SUBTITLE] The Role of Hematologists in Direct Decision Support [SUBSECTION] As shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nAs shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\n[SUBTITLE] Role of Hematologists in Multidisciplinary Collaboration and Providing\nDecision Support [SUBSECTION] We identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.\nWe identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.\n[SUBTITLE] Expected Roles of Other Professionals in Providing Decision Support [SUBSECTION] We extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11).\nWe extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11)."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Participant Characteristics","The Role of Hematologists in Direct Decision Support","Preparing Patients Before Informed Consent (IC)","Selecting the Information To Convey","Choosing the Way In Which To Communicate Information","Respect for The Intentions Of Patients And Families","Directing decision-making and considering fairness","Considering the Emotional Aspects Of The Patient And Family","Providing support after discussing treatments","Role of Hematologists in Multidisciplinary Collaboration and Providing\nDecision Support","Expected Roles of Other Professionals in Providing Decision Support"],"string":"[\n \"Participant Characteristics\",\n \"The Role of Hematologists in Direct Decision Support\",\n \"Preparing Patients Before Informed Consent (IC)\",\n \"Selecting the Information To Convey\",\n \"Choosing the Way In Which To Communicate Information\",\n \"Respect for The Intentions Of Patients And Families\",\n \"Directing decision-making and considering fairness\",\n \"Considering the Emotional Aspects Of The Patient And Family\",\n \"Providing support after discussing treatments\",\n \"Role of Hematologists in Multidisciplinary Collaboration and Providing\\nDecision Support\",\n \"Expected Roles of Other Professionals in Providing Decision Support\"\n]"},"other_sections_texts":{"kind":"list like","value":["We conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).","As shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.","This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.","This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.","This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.","This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.","This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).","This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.","Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.","We identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.","We extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11)."],"string":"[\n \"We conducted semi-structured interviews involving 11 hematologists. The average\\nparticipant age was 44 years (range: 35-54 years), whereas the average\\nspecialization experience was 16 years (range: 7-26 years). Additional details\\nare listed in Table\\n1. The average length of the interviews was 27 minutes (range: 17-46\\nminutes).Table\\n1.Characteristics of doctors and interview\\ndetails.HematologistsTotaln\\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\\n(7-26)*∗Mean(Range).\\nCharacteristics of doctors and interview\\ndetails.\\n∗Mean(Range).\",\n \"As shown in Table 2,\\nwe identified 7 categories related to hematologists’ roles in providing direct\\ndecision support to patients: (1) preparing patients before informed consent,\\n(2) selecting the information to convey, (3) choosing a method for conveying\\nthis information, (4) respecting the intentions of patients and their families,\\n(5) directing decision-making and considering fairness, (6) considering the\\nemotional aspects of patients and their families, and (7) providing support\\nafter discussing treatment options.Table 2.Role of hematologists in providing\\ndirect decision support (n =\\n11).CategorySubcategoryCoden1Preparing patients before\\nIC*(1)Gathering preliminary\\ninformation for explaining issues to patientsObtain information from family members to\\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\\nfamily background1Talk to the family about\\nwhat you will tell the patient1(2)Encouraging family\\nmembers to participate in decision-makingSet up a place to provide information to\\npatients, including family members3Discuss\\nlife-prolonging treatments with the patient’s\\nfamily1(3)Creating an environment for concentrating on\\nICDiscuss in a private\\nroom2Take sufficient time1Leave the\\nPHS*2 with\\nthe nurse so that the discussion is not\\ndisturbed1(4)Providing advance information on the risk\\nof relapseWhen the patient reaches\\ncomplete remission, explain the risk of relapse and\\nprepare the patient for relapse12Selecting the\\ninformation to convey(1)Choice of\\ninformation to provide to patientsIn many cases, do not give the patient a\\nprognosis9Depending on the\\npatient’s medical condition, I (doctor) may not tell the\\ntruth5Modify the message according to the\\npatient’s life stage3(2)Providing patients and families\\nwith information to help them make treatment\\ndecisionsCommunicate the\\nadvantages and disadvantages of treatment\\noptions9Tell patients and their families the\\ninformation they need to make treatment\\ndecisions6Inform patients of all\\ntreatment options4Communicate not only the\\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\\nhelp them understand the patients' current\\nsituationInform family members\\nabout the possibility of sudden changes in the medical\\ncondition and specific prognosis9Provide\\npatients with information related to what they imagine\\nwill happen in the future5Provide patients with\\ninformation that will help them understand their current\\nmedical condition4Provide patients with a\\nvague idea of their prognosis3Communicate the\\nprognosis to patients using concrete numbers23Choosing\\nhow to communicate information(1)Explanations based on the patients' stage of\\nreadinessJudge what I (doctor) say\\nand how I (doctor) say it based on long-standing\\nrelationship with the patients5Provide\\ninformation based on the patient’s reaction and\\nacceptance of their medical condition5(2)Respect\\nfor the preferences of patients and their\\nfamiliesProvide easy-to-understand\\nexplanations without using technical terms5Provide explanations so that patients/families do not\\ndevelop misconceptions4Convey information\\ncarefully and clearly3Provide a written\\nexplanation2Divide bad news into\\ncomponents14Respect for the intentions of\\npatients and families(1)Respect for the\\nwill of patients and familiesRespect and support the patient’s will and desired\\ntreatment6If the patient’s\\ncondition is severe, respect the views of the family and\\nprimary caregiver2(2)Setting goals in accordance\\nwith the preferences of the patients and\\nfamiliesListen to the needs of\\npatients and their families and determine actionable\\ntreatment goals3Explore the intentions of\\npatients and their families through\\nconversations1If there is no\\ndifference between the advantages and disadvantages of a\\ntreatment choice, the choice is purely made by the\\npatient15Directing decision-making and considering\\nfairness(1)Considering the direction and\\nfairness of treatment options recommended by\\ndoctorsExplain treatment options\\nwith some direction while also providing options that\\nthe doctor finds beneficial8In some cases,\\noffer personal opinion as a doctor2Dare to treat\\nthe patient with paternalism2(2)Consideration for\\npatients to make impartial decisionsCommunicate with the knowledge that the\\ndoctor’s manner of speaking influences the patients’\\ndecisions5Explain the opinions of\\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\\nfamily(1)Explaining with consideration\\nfor emotionsTalk in a way that\\nallows the patient to have hope5Convey bad news\\naccording to the protocol1Consider that patients\\ndo not have to listen to explanations while always\\nlooking at bad information1Divulge bad news in\\nstages1(2)Showing a supportive attitude to the\\npatientApproach the patient with a\\nnon-abandoning attitude2Show that the team\\nsupports the patient1(3)Attitude when facing patients\\nand their familiesInteract with\\npatients on an equal footing2Talk while\\nlooking into the patient’s eyes1Interact with\\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\\ntreatmentsSupport fluctuations in\\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\nRole of hematologists in providing\\ndirect decision support (n =\\n11).\\n*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\nThis category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\nThis category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\nSeveral participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\",\n \"This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\",\n \"This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\",\n \"This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\",\n \"This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\",\n \"This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\",\n \"This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\",\n \"Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\",\n \"We identified the following 5 subcategories related to the second main study\\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\\ncommunicating with other professionals, (2) gathering information from them, (3)\\nproviding information to them, (4) managing the entire medical team, and (5)\\nencouraging nurses to actively participate with patients throughout the\\ndecision-making process (Table 3). More than half of the participants recognized the\\nhematologist’s role in communicating with other professionals and obtaining\\ninformation about the patient (eg, the patient’s background and how to help)\\nthroughout this process. Others said that they actively encourage nurses to\\nsupport the patients’ decisions and provide additional information during the\\ndiscussions or explain the prognoses and future treatment plans to other\\nprofessionals.Table\\n3.The role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\\nprofessionals, such as nursesMake opportunities to hold discussions with\\nmultiple professionals2Gathering\\ninformation from other professionalsObtain information from nurses about family\\nbackground and family supportObtain information about the patients’\\nthoughts and intentions from other\\nprofessionalsObtain\\ninformation about the patients’ social background from\\nother professionalsObtain\\ninformation about the patient’s ADL status from other\\nprofessionals3Providing\\ninformation to other professionalsExplain the patient’s prognosis and future treatment\\nplans to other professionalsExplain the concerns of the patient’s\\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\\nhematologists make comprehensive decisions based on\\ninformation from other professionalsCreate a support system for\\npatients5Encouraging nurses to actively\\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\\npatients’ decision-makingConfirm whether there is any additional information or\\nthoughts to convey to nurses during discussions about\\ntreatment decisions with the\\npatientADL\\n= Activities of Daily\\nLiving.\\nThe role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).\\nADL\\n= Activities of Daily\\nLiving.\",\n \"We extracted the following 4 categories related to the roles that hematologists\\nexpected other professionals to fulfill: (1) sharing patient information and\\ncare routines, (2) sharing information from treatment discussions, (3) providing\\nemotional support to patients and their families, and (4) assembling a medical\\nteam that understands hematological malignancies (Table 4). Nearly half of the\\nparticipants indicated that they would like other health care providers to\\nconvey the patients’ thoughts and intentions that they knew to the doctors\\ninvolved. They also expected other professionals to share information regarding\\ntheir patients and the way in which they cared for their patients.Table 4.Expected\\nroles of other professionals in decision support (n =\\n11).SubcategoryCodenSharing patient information and care\\nroutinesListen to the patients’\\nthoughts and wishes and communicate them to the\\ndoctor5Suggest knowledge that doctors may not know\\nand explain how to care for patients in the right\\nway5Point out the problems patients have with treatment\\ndecisions2Sharing information from the\\ntreatment discussionAttend\\ntreatment discussions and understand the patients’\\nfeelings and situation together4Read records of the\\nprocess of informed consent and share the\\ninformation1Providing patients and their families with\\nemotional supportSupport the\\nemotions of patients and their families4Assembling\\na medical team that understands hematological\\nmalignancyWork to ensure that the\\nmedical team understands hematological\\nmalignancy1\\nExpected\\nroles of other professionals in decision support (n =\\n11).\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Results","Participant Characteristics","The Role of Hematologists in Direct Decision Support","Preparing Patients Before Informed Consent (IC)","Selecting the Information To Convey","Choosing the Way In Which To Communicate Information","Respect for The Intentions Of Patients And Families","Directing decision-making and considering fairness","Considering the Emotional Aspects Of The Patient And Family","Providing support after discussing treatments","Role of Hematologists in Multidisciplinary Collaboration and Providing\nDecision Support","Expected Roles of Other Professionals in Providing Decision Support","Discussion","Supplemental Material"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Results\",\n \"Participant Characteristics\",\n \"The Role of Hematologists in Direct Decision Support\",\n \"Preparing Patients Before Informed Consent (IC)\",\n \"Selecting the Information To Convey\",\n \"Choosing the Way In Which To Communicate Information\",\n \"Respect for The Intentions Of Patients And Families\",\n \"Directing decision-making and considering fairness\",\n \"Considering the Emotional Aspects Of The Patient And Family\",\n \"Providing support after discussing treatments\",\n \"Role of Hematologists in Multidisciplinary Collaboration and Providing\\nDecision Support\",\n \"Expected Roles of Other Professionals in Providing Decision Support\",\n \"Discussion\",\n \"Supplemental Material\"\n]"},"all_sections_texts":{"kind":"list like","value":["The prognosis and illness trajectory of leukemia and lymphoma vary greatly depending\non the degree of progression and subtype of the disease. However, irrespective of\nthe illness trajectory, patients face the same issues and debates related to\ndecision-making, such as how far to go with aggressive treatment during the relapse\nor treatment-resistant period and the choice of treatment and place of\ncare.1-3 Even in relapse cases, such\npatients may have the option to receive high-dose chemotherapy treatments or\nhematopoietic cell transplantations aimed at remission and achieving long-term\nsurvival. Such treatment carries significant risks, including complications such as\nsevere infections and bleeding due to bone marrow suppression and treatment-related\ndeath, and may result in transition from invasive treatment aimed at cure to\nend-of-life (EOL). Even when the disease is refractory to treatment, anticancer\ntherapy aimed at disease control is often administered until the end of life. In\nsuch contexts, it is very difficult for patients and their families to make\ndecisions about these complex treatment strategies and the high risks involved in\nthe context of their own life values.4-6\nThe American Society of Clinical Oncology (ASCO) has published consensus guidelines\nfor clinician-patient communication, specifically recommending that clinicians\nprovide information that is congruent with their patients’ interests and intentions\nwhile jointly considering their goals for treatment.7 It is also notable that\nhematologists’ communication skills have been receiving increasing attention in\nhematology clinical practice in recent years. An ethnographic study of\nhematologists' communication patterns when communicating bad news revealed 4\ncharacteristics: (1) technical-defensive patterns, (2) an authoritative pattern, (3)\na relational-recursive pattern, and (4) a compassionate sharing pattern, and the\npaper also suggested that hematologists were not good at expressing caring and\nempathy.8 In contrast, a qualitative study examining the needs of patients\nwith hematological malignancies when being told bad news reported identifying\nemotional support from their healthcare providers and information needs about future\nsupport systems. It was also noted that healthcare-team involvement is necessary to\nmeet these patients' needs and support their decision-making.9\nTherefore, this study sought to clarify the following: (a) the way in which\nhematologists perceive the roles of doctors in providing decision-making support to\npatients suffering from relapsed and refractory leukemia as well as lymphoma, (b)\nthe expected roles of hematologists in supporting decision-making via\nmultidisciplinary collaboration, and (c) the roles hematologists expect other\nhealthcare providers to fulfill.","The purpose of this study was to determine hematologists’ role in decision support\nfor patients with leukemia and lymphoma. Physicians at institutions with hematology\nwards and hematology outpatient clinics where chemotherapy and bone marrow\ntransplantation are performed were included. The eligibility criteria for physicians\nincluded being Board Certified Hematologists and having at least 5 years of\nexperience. Hematologists meeting these criteria were considered to have experience\nseeing patients with relapsed or refractory leukemia or lymphoma. Subjects were\nselected using opportunistic sampling, which is used in selecting those with common\ncharacteristics and the potential to provide a rich, appropriate, and diverse\nresearch agenda. Eligible physicians were asked to recommend a qualified physician\nwho met the inclusion criteria to the head of the hematology department of a\nhospital in Tokyo. The recommended physicians were informed orally and in writing\nabout the purpose and methods of the study and the voluntary participation clause.\nAlthough this is a qualitative study and it is difficult to theoretically calculate\nthe sample size, we chose a sample size of 10 to 15 participants based on previous\nstudies9-11 using the\nsame content analysis method as in this study. The acceptance rate of recommended\nphysicians was 100%.\nVerbal and written explanations of the research objectives and procedures were\nprovided to all participants, and written consent was obtained. This study was\napproved by the Institutional Review Board of University of Tokyo (Approval No.\n3443). The study was initiated at the University of Tokyo and completed at Kyoto\nUniversity.\nTo collect the data, we conducted in-depth interviews to identify the attributes and\nroles of doctors as well as the expected roles of nurses in providing\ndecision-making support to patients with relapsed or refractory leukemia and\nlymphoma. Specifically, we scheduled face-to-face semi-structured interviews in\nprivate rooms between August and December 2011. All the interviews conducted were\naudio recorded and fully transcribed by 1 researcher (a research nurse), who also\ntook detailed field notes during each interview. Physicians were first asked for\ntheir demographic details, including age, the type of facility where they worked,\nand the number of years they had worked in their area of specialization. We then\nasked them about their roles in treating patients suffering from relapsed or\nrefractory leukemia and lymphoma.\nTo analyze the data, interview data were transcribed verbatim from the audio\nrecordings. Interview data were then numbered for each participant, separated from\npersonal information, and anonymized. A content analysis was then conducted. First,\n1 researcher extracted all statements related to the 3 study topics from each\ntranscript (eg, the role of hematologists, the role of hematologists in\nmultidisciplinary collaboration, and the expectations of other professionals in the\ncontext of providing decision-making support). We then carefully conceptualized the\ntext and categorized it into content areas through content analysis using the Klaus\nKrippendorff12 method. The units, codes, and categories were decided\nthrough consensus. Next, another experienced qualitative researcher read the\noriginal interviews and classified them into categories and subcategories. The coder\nread the original interview text and independently determined the subcategory into\nwhich the content fell. The rate of agreement between the researchers and coders was\n91%. Finally, the coder and the researcher discussed the units they disagreed on\nuntil they reached a consensus.","[SUBTITLE] Participant Characteristics [SUBSECTION] We conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).\nWe conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).\n[SUBTITLE] The Role of Hematologists in Direct Decision Support [SUBSECTION] As shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nAs shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\n[SUBTITLE] Role of Hematologists in Multidisciplinary Collaboration and Providing\nDecision Support [SUBSECTION] We identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.\nWe identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.\n[SUBTITLE] Expected Roles of Other Professionals in Providing Decision Support [SUBSECTION] We extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11).\nWe extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11).","We conducted semi-structured interviews involving 11 hematologists. The average\nparticipant age was 44 years (range: 35-54 years), whereas the average\nspecialization experience was 16 years (range: 7-26 years). Additional details\nare listed in Table\n1. The average length of the interviews was 27 minutes (range: 17-46\nminutes).Table\n1.Characteristics of doctors and interview\ndetails.HematologistsTotaln\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\n(7-26)*∗Mean(Range).\nCharacteristics of doctors and interview\ndetails.\n∗Mean(Range).","As shown in Table 2,\nwe identified 7 categories related to hematologists’ roles in providing direct\ndecision support to patients: (1) preparing patients before informed consent,\n(2) selecting the information to convey, (3) choosing a method for conveying\nthis information, (4) respecting the intentions of patients and their families,\n(5) directing decision-making and considering fairness, (6) considering the\nemotional aspects of patients and their families, and (7) providing support\nafter discussing treatment options.Table 2.Role of hematologists in providing\ndirect decision support (n =\n11).CategorySubcategoryCoden1Preparing patients before\nIC*(1)Gathering preliminary\ninformation for explaining issues to patientsObtain information from family members to\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\nfamily background1Talk to the family about\nwhat you will tell the patient1(2)Encouraging family\nmembers to participate in decision-makingSet up a place to provide information to\npatients, including family members3Discuss\nlife-prolonging treatments with the patient’s\nfamily1(3)Creating an environment for concentrating on\nICDiscuss in a private\nroom2Take sufficient time1Leave the\nPHS*2 with\nthe nurse so that the discussion is not\ndisturbed1(4)Providing advance information on the risk\nof relapseWhen the patient reaches\ncomplete remission, explain the risk of relapse and\nprepare the patient for relapse12Selecting the\ninformation to convey(1)Choice of\ninformation to provide to patientsIn many cases, do not give the patient a\nprognosis9Depending on the\npatient’s medical condition, I (doctor) may not tell the\ntruth5Modify the message according to the\npatient’s life stage3(2)Providing patients and families\nwith information to help them make treatment\ndecisionsCommunicate the\nadvantages and disadvantages of treatment\noptions9Tell patients and their families the\ninformation they need to make treatment\ndecisions6Inform patients of all\ntreatment options4Communicate not only the\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\nhelp them understand the patients' current\nsituationInform family members\nabout the possibility of sudden changes in the medical\ncondition and specific prognosis9Provide\npatients with information related to what they imagine\nwill happen in the future5Provide patients with\ninformation that will help them understand their current\nmedical condition4Provide patients with a\nvague idea of their prognosis3Communicate the\nprognosis to patients using concrete numbers23Choosing\nhow to communicate information(1)Explanations based on the patients' stage of\nreadinessJudge what I (doctor) say\nand how I (doctor) say it based on long-standing\nrelationship with the patients5Provide\ninformation based on the patient’s reaction and\nacceptance of their medical condition5(2)Respect\nfor the preferences of patients and their\nfamiliesProvide easy-to-understand\nexplanations without using technical terms5Provide explanations so that patients/families do not\ndevelop misconceptions4Convey information\ncarefully and clearly3Provide a written\nexplanation2Divide bad news into\ncomponents14Respect for the intentions of\npatients and families(1)Respect for the\nwill of patients and familiesRespect and support the patient’s will and desired\ntreatment6If the patient’s\ncondition is severe, respect the views of the family and\nprimary caregiver2(2)Setting goals in accordance\nwith the preferences of the patients and\nfamiliesListen to the needs of\npatients and their families and determine actionable\ntreatment goals3Explore the intentions of\npatients and their families through\nconversations1If there is no\ndifference between the advantages and disadvantages of a\ntreatment choice, the choice is purely made by the\npatient15Directing decision-making and considering\nfairness(1)Considering the direction and\nfairness of treatment options recommended by\ndoctorsExplain treatment options\nwith some direction while also providing options that\nthe doctor finds beneficial8In some cases,\noffer personal opinion as a doctor2Dare to treat\nthe patient with paternalism2(2)Consideration for\npatients to make impartial decisionsCommunicate with the knowledge that the\ndoctor’s manner of speaking influences the patients’\ndecisions5Explain the opinions of\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\nfamily(1)Explaining with consideration\nfor emotionsTalk in a way that\nallows the patient to have hope5Convey bad news\naccording to the protocol1Consider that patients\ndo not have to listen to explanations while always\nlooking at bad information1Divulge bad news in\nstages1(2)Showing a supportive attitude to the\npatientApproach the patient with a\nnon-abandoning attitude2Show that the team\nsupports the patient1(3)Attitude when facing patients\nand their familiesInteract with\npatients on an equal footing2Talk while\nlooking into the patient’s eyes1Interact with\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\ntreatmentsSupport fluctuations in\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\nRole of hematologists in providing\ndirect decision support (n =\n11).\n*1IC\n= informed consent; *2PHS = Personal Handyphone\nSystem (A phone used by doctors to communicate with staff in the\nhospital).\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\nThis category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\nThis category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.\nSeveral participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.","This category comprised 4 subcategories. Several participants talked about\npreparing patients and families for IC discussions, including “Gathering\npreliminary information for explaining issues to patients” and “Encouraging\nfamily members to participate in decision-making.” As such, physicians were\nexpected to prepare patients before IC discussions. For instance, they would\nprovide patients with information regarding the risk of relapse.","This category comprised 3 subcategories. In subcategory (1) “Choice of\ninformation to provide to patients,” more than half of the participants said\nthat they did not provide their patients with prognoses. More importantly,\nmore than half said they may not even tell their patients the truth,\ndepending on the severity of the illness, to avoid a loss of hope; they were\nvery selective about the information they chose to provide to their\npatients.\nIn subcategory (2) “Providing patients and families with information to help\nthem make treatment decisions,” more than half of the participants\nidentified “Communicating the advantages and disadvantages of treatment\noptions” and “Telling patients and their families the information they need\nto make treatment decisions.” They provided information they believed would\nhelp patients and their families fully understand all the options before\nmaking a final decision.\nIn subcategory (3), “Providing patients and families with information to help\nthem understand the patients’ current situation,” a large number said they\nwould inform family members about the possibility of sudden changes and\nspecific prognoses, with more than half saying they would explain the issues\nso that patients could visualize what may happen to them in the future.\nGenerally, participants provided information to help patients and their\nfamilies clearly understand their current situations.","This category comprised 2 subcategories. Nearly half of participants\nexpressed the ideas of “I determine what I say and how I say it based on our\nlong-standing relationship with the patients” and “Providing information\nbased on the patient’s reaction and acceptance of their medical condition.”\nThey carefully considered what they discussed and how they discussed it,\nespecially considering their long-standing association with their patients.\nIn this regard, they conveyed information based on how they thought patients\nwould react and their level of acceptance. They also provided explanations\naccording to what they believed patients intended and thought depending on\ntheir respective stages of readiness. Moreover, nearly half said that they\nwould explain issues without using jargon, thereby ensuring that the\ninformation provided was easy to understand. Essentially, participants\nremained cognizant of the need to communicate both correctly and\nappropriately with patients and their families.","This category comprised 2 subcategories. In subcategory (1) “Respect for the\nwill of patients and families,” more than half of the participants said they\nhelped patients and families make decisions by supporting and respecting\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\naccordance with the preferences of the patients and families,” several\nparticipants said that they helped patients and families set feasible goals\nafter listening to their wishes.","This category comprised 2 subcategories. In subcategory (1) “Considering the\ndirection and fairness of treatment options recommended by doctors,” more\nthan half of the participants said they explain treatment options with some\ndirection while also providing options they find most beneficial. They\ntalked about giving weight to treatment options and facilitating the\ndecision-making process. Furthermore, they attempted to present information\nin a fair manner, being aware that the way they talk can affect the way\npatients perceive their treatment, as shown in subcategory (2).","This category comprised 3 subcategories. In subcategory (1)“explaining with\nconsideration for emotions,” nearly half of the participants said they would\nattempt to instill hope among patients, and hence, provide explanations\nconsidering the related emotional aspects. In subcategory (2) “showing a\nsupportive attitude to the patient”, several participants also expressed\nadopting a supportive stance to reassure patients that they were not\nabandoning them or their families. In subcategory (3) “attitude when facing\npatients and their families”, several participants talked about the doctor’s\nattitude toward patients, such as speaking with patients and families as\nequals and looking them in the eye.","Several participants said they informed patients about bad news and future\ntreatments, after which they provided emotional support in consideration of\nany emotional variabilities.","We identified the following 5 subcategories related to the second main study\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\ncommunicating with other professionals, (2) gathering information from them, (3)\nproviding information to them, (4) managing the entire medical team, and (5)\nencouraging nurses to actively participate with patients throughout the\ndecision-making process (Table 3). More than half of the participants recognized the\nhematologist’s role in communicating with other professionals and obtaining\ninformation about the patient (eg, the patient’s background and how to help)\nthroughout this process. Others said that they actively encourage nurses to\nsupport the patients’ decisions and provide additional information during the\ndiscussions or explain the prognoses and future treatment plans to other\nprofessionals.Table\n3.The role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\nprofessionals, such as nursesMake opportunities to hold discussions with\nmultiple professionals2Gathering\ninformation from other professionalsObtain information from nurses about family\nbackground and family supportObtain information about the patients’\nthoughts and intentions from other\nprofessionalsObtain\ninformation about the patients’ social background from\nother professionalsObtain\ninformation about the patient’s ADL status from other\nprofessionals3Providing\ninformation to other professionalsExplain the patient’s prognosis and future treatment\nplans to other professionalsExplain the concerns of the patient’s\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\nhematologists make comprehensive decisions based on\ninformation from other professionalsCreate a support system for\npatients5Encouraging nurses to actively\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\npatients’ decision-makingConfirm whether there is any additional information or\nthoughts to convey to nurses during discussions about\ntreatment decisions with the\npatientADL\n= Activities of Daily\nLiving.\nThe role of hematologists in multidisciplinary\ncollaboration and decision support (n =\n11).\nADL\n= Activities of Daily\nLiving.","We extracted the following 4 categories related to the roles that hematologists\nexpected other professionals to fulfill: (1) sharing patient information and\ncare routines, (2) sharing information from treatment discussions, (3) providing\nemotional support to patients and their families, and (4) assembling a medical\nteam that understands hematological malignancies (Table 4). Nearly half of the\nparticipants indicated that they would like other health care providers to\nconvey the patients’ thoughts and intentions that they knew to the doctors\ninvolved. They also expected other professionals to share information regarding\ntheir patients and the way in which they cared for their patients.Table 4.Expected\nroles of other professionals in decision support (n =\n11).SubcategoryCodenSharing patient information and care\nroutinesListen to the patients’\nthoughts and wishes and communicate them to the\ndoctor5Suggest knowledge that doctors may not know\nand explain how to care for patients in the right\nway5Point out the problems patients have with treatment\ndecisions2Sharing information from the\ntreatment discussionAttend\ntreatment discussions and understand the patients’\nfeelings and situation together4Read records of the\nprocess of informed consent and share the\ninformation1Providing patients and their families with\nemotional supportSupport the\nemotions of patients and their families4Assembling\na medical team that understands hematological\nmalignancyWork to ensure that the\nmedical team understands hematological\nmalignancy1\nExpected\nroles of other professionals in decision support (n =\n11).","In this study, hematologists were interviewed to clarify their role and expectations\nof other professionals. The categories of hematologists’ direct role in\ndecision-making were extracted, including preparation and consideration of the\nsituation in which information about decision-making is communicated and emotional\nsupport after the information is communicated.\nHematologists presented treatment options based on a variety of factors, including\nthe patient’s goals, emotions, the preparedness for the disease, and the severity of\nthe disease. During this process, the hematologists worked with other professionals\nto understand specific information about the patient and family, including their\nbackground, goals, and intentions. They also said that they would decide what\ninformation to convey and carefully explain the treatment and medical condition,\nwithout using jargon, so that the patient and family could easily understand. ASCO\nguidelines recommend that physicians provide information that takes into account\nindividual patient interests and preferences.7 Although the present results\nare data obtained before these guidelines were developed, the hematologists who\nparticipated in this study were taking the actions recommended in the guidelines.\nThe present results reveal more specific behaviors of hematologists with regard to\nwhat is recommended in the guidelines. For patients facing complex and risky\ntreatment choices, it is important for hematologist to select information according\nto the patient’s wishes and to convey information in easy-to-understand terms.\nHematologists were expected other professionalsto pay attention to the emotional\naspects of the patient and family to minimize their anxiety. In addition, they\nexpected nurses and other professionals to listen to the patient and family’s wishes\nand thoughts about treatment, share them with the physician, and serve as emotional\nsupport for the patient and family. Moreover, hematologists were expected to\nunderstand the patient’s wishes and preferences, explain complex and\ndifficult-to-understand treatments and medical conditions, and present options that\nwere consistent with the patient’s wishes. ASCO guidelines also emphasize the need\nto form a rapport with patients and their families and to assess and support coping\nneeds.7 The hematologists who participated in this study wanted to\nimplement this in collaboration with other professionals. The wishes and intentions\nof patients and their families change as the disease progresses or symptoms worsen.\nTherefore, it is thought that by having physicians and other professions work\ntogether to understand the wishes of patients/families at that time, decision-making\nsupport in accordance with the wishes of patients/families will become possible.\nLeukemia and lymphoma may be treated through chemotherapy, even in cases of relapse\nand during the refractory phase. This makes it difficult to determine when to\ntransition from curative treatments to those more focused on quality of life\n(QOL).13,14 Previous studies have reported that hematologists are reluctant\nto discuss EOL issues while primary disease treatments are still possible.15,16 A number of\nthe hematologists who participated in the study had a variety of ways of explaining\nthe prognosis to patients and their families. For example, in some cases, they\ninformed the family about the possibility of a sudden change in prognosis, including\nspecific figures, while in other cases, depending on the severity of the disease and\nthe stage of life, they did not inform them of the prognosis at all or withheld the\nfact. Furthermore, more than half of the respondents stated that they would explain\nthe treatment that they thought would be most beneficial to the patient.\nA number of the hematologists who participated in this study also had a long history\nof working with patients to understand their individual backgrounds, and what they\ntalked about and how they communicate, while also having a direction that the\nhematologists thought was suitable, and gave the patients some direction as they\nspoke. This suggested that hematologists attempt to tailor their decision-making\nsupport to the needs of their patients. On the other hand, hematologists infer\npatients' needs based on their previous experience, which runs the risk of making\nsuggestions that are not in line with the patients’ original needs. Previous studies\nhave reported that hematologic oncology patients take a passive role in the\ntreatment decision-making process due to the complexity of their treatment and\ndisease17 and expect their physicians to take a paternalistic\nrole.16 When hematologists provide some treatment direction and\npaternalistic response to hematologic oncology patients with these characteristics,\npatients may not fully understand the advantages and disadvantages of treatment and\nmay not make an informed choice. If treatment is responded to, the patient’s quality\nof life improves, but if treatment is not effective, the patient may die in the\nhospital during active treatment. In addition, patients may not be able to live the\nlife they want, with increased risk of infection and prolonged\nhospitalization.1 Hematologists need to have frank discussions with patients,\nkeeping in mind the direction they think is best for the patient, and working with\nother professionals to understand the patient’s wishes.\nHematologists who participated in this study indicated that the role they expected of\nmultidisciplinary professionals was to inform them about patient thoughts,\ninformation, and methods of care that the physicians were unaware of. They also\nexpected other professions to play a role in providing emotional support to patients\nand their families. In a previous study of oncologists, including hematologists,\nlack of information about the patient was the most frequently cited factor hindering\npatient involvement in decision-making.18 Physicians also recognized\nthat having a third party present during decision-making discussions supported\npatient decision-making and facilitated involvement and reflection on treatment\ndecisions.18 In that previous study, physicians emphasized the\nimportance of obtaining information about the patient and the presence of a third\nparty in the decision-making situation, which was similar to the findings of the\npresent study. This confirms the importance of a multidisciplinary approach to\nfacilitate decision-making in accordance with the patient’s wishes. In making\ntreatment decisions, it is necessary to organize information not only from a medical\nperspective, but also based on the patient’s priorities and values. Obtaining such\ninformation is difficult for hematologists alone, and multidisciplinary\ncollaboration is important. By understanding the patient’s needs and organizing the\ninformation necessary for decision-making, it is expected that the discussion\nbetween the attending hematologist and the patient will be deepened.9\nFinally, it is important to note that the survey for this study was conducted in 2011\nand there was not much collaboration with the palliative care team,15 the\nparticipants in this study did not mention the role they expected to play regarding\npalliative care physicians. Recently, however, the effects of early palliative care\nhave been reported in clinical hematologic oncology. A study evaluating the effects\nof early introduction of palliative care and continuous psychological support by a\nmultidisciplinary team in patients with acute leukemia reported a reduction in\nsymptoms due to acute stress reactions compared to the usual care group.19 It has also\nbeen reported that patients undergoing hematopoietic stem cell transplantation who\nreceived regular twice-weekly visits by a palliative care physician had a higher QOL\nat 3 months of hospitalization compared to the group that received usual\ncare.20 In patients with leukemia and lymphoma, where the potential for\nrapid change is high, it is expected that hematologists and palliative care\nspecialists will support patients and begin discussing long-term goals early, in\ncase treatment goals suddenly change.21\nThis study had some limitations. The first is that the data collection and analysis\nwere conducted between 2011 and 2012. Therefore, because treatment outcomes for\nhematologic tumors have improved compared to a decade ago, the data cannot be\ndirectly applied to decision support for patients with such tumors and their\nfamilies today. However, the patients with treatment-resistant hematologic tumors\nhave not changed significantly in their characteristics or treatment\nenvironment,1,22 and it is significant to clarify the hematologist’s role in\ndecision support. Second, the use of the opportunistic sampling makes it difficult\nto generalize the findings. In this study, we collected the opinions of\nhematologists at hospitals where hematologic oncology patients are mainly treated,\nsuch as cancer hospitals and university hospitals in Japan, and we believe that we\nwere able to collect diverse opinions regarding the role of decision support as\nperceived by hematologists. Based on the present study, research is needed to\nclarify the decision support roles of other professionals who care for patients with\nleukemia and lymphoma, including nurses. This would make it possible to construct a\nmodel of shared decision-making support for hematologic oncology patients through\nmultidisciplinary collaboration.","Click here for additional data file.\nSupplemental Material for Qualitative Analysis of the Roles of Physicians and\nNurses in Providing Decision Support to Patients With Relapsed or Refractory\nLeukemia and Lymphoma by Miharu Morikawa, and Yuki Shirai in Cancer Control"],"string":"[\n \"The prognosis and illness trajectory of leukemia and lymphoma vary greatly depending\\non the degree of progression and subtype of the disease. However, irrespective of\\nthe illness trajectory, patients face the same issues and debates related to\\ndecision-making, such as how far to go with aggressive treatment during the relapse\\nor treatment-resistant period and the choice of treatment and place of\\ncare.1-3 Even in relapse cases, such\\npatients may have the option to receive high-dose chemotherapy treatments or\\nhematopoietic cell transplantations aimed at remission and achieving long-term\\nsurvival. Such treatment carries significant risks, including complications such as\\nsevere infections and bleeding due to bone marrow suppression and treatment-related\\ndeath, and may result in transition from invasive treatment aimed at cure to\\nend-of-life (EOL). Even when the disease is refractory to treatment, anticancer\\ntherapy aimed at disease control is often administered until the end of life. In\\nsuch contexts, it is very difficult for patients and their families to make\\ndecisions about these complex treatment strategies and the high risks involved in\\nthe context of their own life values.4-6\\nThe American Society of Clinical Oncology (ASCO) has published consensus guidelines\\nfor clinician-patient communication, specifically recommending that clinicians\\nprovide information that is congruent with their patients’ interests and intentions\\nwhile jointly considering their goals for treatment.7 It is also notable that\\nhematologists’ communication skills have been receiving increasing attention in\\nhematology clinical practice in recent years. An ethnographic study of\\nhematologists' communication patterns when communicating bad news revealed 4\\ncharacteristics: (1) technical-defensive patterns, (2) an authoritative pattern, (3)\\na relational-recursive pattern, and (4) a compassionate sharing pattern, and the\\npaper also suggested that hematologists were not good at expressing caring and\\nempathy.8 In contrast, a qualitative study examining the needs of patients\\nwith hematological malignancies when being told bad news reported identifying\\nemotional support from their healthcare providers and information needs about future\\nsupport systems. It was also noted that healthcare-team involvement is necessary to\\nmeet these patients' needs and support their decision-making.9\\nTherefore, this study sought to clarify the following: (a) the way in which\\nhematologists perceive the roles of doctors in providing decision-making support to\\npatients suffering from relapsed and refractory leukemia as well as lymphoma, (b)\\nthe expected roles of hematologists in supporting decision-making via\\nmultidisciplinary collaboration, and (c) the roles hematologists expect other\\nhealthcare providers to fulfill.\",\n \"The purpose of this study was to determine hematologists’ role in decision support\\nfor patients with leukemia and lymphoma. Physicians at institutions with hematology\\nwards and hematology outpatient clinics where chemotherapy and bone marrow\\ntransplantation are performed were included. The eligibility criteria for physicians\\nincluded being Board Certified Hematologists and having at least 5 years of\\nexperience. Hematologists meeting these criteria were considered to have experience\\nseeing patients with relapsed or refractory leukemia or lymphoma. Subjects were\\nselected using opportunistic sampling, which is used in selecting those with common\\ncharacteristics and the potential to provide a rich, appropriate, and diverse\\nresearch agenda. Eligible physicians were asked to recommend a qualified physician\\nwho met the inclusion criteria to the head of the hematology department of a\\nhospital in Tokyo. The recommended physicians were informed orally and in writing\\nabout the purpose and methods of the study and the voluntary participation clause.\\nAlthough this is a qualitative study and it is difficult to theoretically calculate\\nthe sample size, we chose a sample size of 10 to 15 participants based on previous\\nstudies9-11 using the\\nsame content analysis method as in this study. The acceptance rate of recommended\\nphysicians was 100%.\\nVerbal and written explanations of the research objectives and procedures were\\nprovided to all participants, and written consent was obtained. This study was\\napproved by the Institutional Review Board of University of Tokyo (Approval No.\\n3443). The study was initiated at the University of Tokyo and completed at Kyoto\\nUniversity.\\nTo collect the data, we conducted in-depth interviews to identify the attributes and\\nroles of doctors as well as the expected roles of nurses in providing\\ndecision-making support to patients with relapsed or refractory leukemia and\\nlymphoma. Specifically, we scheduled face-to-face semi-structured interviews in\\nprivate rooms between August and December 2011. All the interviews conducted were\\naudio recorded and fully transcribed by 1 researcher (a research nurse), who also\\ntook detailed field notes during each interview. Physicians were first asked for\\ntheir demographic details, including age, the type of facility where they worked,\\nand the number of years they had worked in their area of specialization. We then\\nasked them about their roles in treating patients suffering from relapsed or\\nrefractory leukemia and lymphoma.\\nTo analyze the data, interview data were transcribed verbatim from the audio\\nrecordings. Interview data were then numbered for each participant, separated from\\npersonal information, and anonymized. A content analysis was then conducted. First,\\n1 researcher extracted all statements related to the 3 study topics from each\\ntranscript (eg, the role of hematologists, the role of hematologists in\\nmultidisciplinary collaboration, and the expectations of other professionals in the\\ncontext of providing decision-making support). We then carefully conceptualized the\\ntext and categorized it into content areas through content analysis using the Klaus\\nKrippendorff12 method. The units, codes, and categories were decided\\nthrough consensus. Next, another experienced qualitative researcher read the\\noriginal interviews and classified them into categories and subcategories. The coder\\nread the original interview text and independently determined the subcategory into\\nwhich the content fell. The rate of agreement between the researchers and coders was\\n91%. Finally, the coder and the researcher discussed the units they disagreed on\\nuntil they reached a consensus.\",\n \"[SUBTITLE] Participant Characteristics [SUBSECTION] We conducted semi-structured interviews involving 11 hematologists. The average\\nparticipant age was 44 years (range: 35-54 years), whereas the average\\nspecialization experience was 16 years (range: 7-26 years). Additional details\\nare listed in Table\\n1. The average length of the interviews was 27 minutes (range: 17-46\\nminutes).Table\\n1.Characteristics of doctors and interview\\ndetails.HematologistsTotaln\\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\\n(7-26)*∗Mean(Range).\\nCharacteristics of doctors and interview\\ndetails.\\n∗Mean(Range).\\nWe conducted semi-structured interviews involving 11 hematologists. The average\\nparticipant age was 44 years (range: 35-54 years), whereas the average\\nspecialization experience was 16 years (range: 7-26 years). Additional details\\nare listed in Table\\n1. The average length of the interviews was 27 minutes (range: 17-46\\nminutes).Table\\n1.Characteristics of doctors and interview\\ndetails.HematologistsTotaln\\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\\n(7-26)*∗Mean(Range).\\nCharacteristics of doctors and interview\\ndetails.\\n∗Mean(Range).\\n[SUBTITLE] The Role of Hematologists in Direct Decision Support [SUBSECTION] As shown in Table 2,\\nwe identified 7 categories related to hematologists’ roles in providing direct\\ndecision support to patients: (1) preparing patients before informed consent,\\n(2) selecting the information to convey, (3) choosing a method for conveying\\nthis information, (4) respecting the intentions of patients and their families,\\n(5) directing decision-making and considering fairness, (6) considering the\\nemotional aspects of patients and their families, and (7) providing support\\nafter discussing treatment options.Table 2.Role of hematologists in providing\\ndirect decision support (n =\\n11).CategorySubcategoryCoden1Preparing patients before\\nIC*(1)Gathering preliminary\\ninformation for explaining issues to patientsObtain information from family members to\\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\\nfamily background1Talk to the family about\\nwhat you will tell the patient1(2)Encouraging family\\nmembers to participate in decision-makingSet up a place to provide information to\\npatients, including family members3Discuss\\nlife-prolonging treatments with the patient’s\\nfamily1(3)Creating an environment for concentrating on\\nICDiscuss in a private\\nroom2Take sufficient time1Leave the\\nPHS*2 with\\nthe nurse so that the discussion is not\\ndisturbed1(4)Providing advance information on the risk\\nof relapseWhen the patient reaches\\ncomplete remission, explain the risk of relapse and\\nprepare the patient for relapse12Selecting the\\ninformation to convey(1)Choice of\\ninformation to provide to patientsIn many cases, do not give the patient a\\nprognosis9Depending on the\\npatient’s medical condition, I (doctor) may not tell the\\ntruth5Modify the message according to the\\npatient’s life stage3(2)Providing patients and families\\nwith information to help them make treatment\\ndecisionsCommunicate the\\nadvantages and disadvantages of treatment\\noptions9Tell patients and their families the\\ninformation they need to make treatment\\ndecisions6Inform patients of all\\ntreatment options4Communicate not only the\\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\\nhelp them understand the patients' current\\nsituationInform family members\\nabout the possibility of sudden changes in the medical\\ncondition and specific prognosis9Provide\\npatients with information related to what they imagine\\nwill happen in the future5Provide patients with\\ninformation that will help them understand their current\\nmedical condition4Provide patients with a\\nvague idea of their prognosis3Communicate the\\nprognosis to patients using concrete numbers23Choosing\\nhow to communicate information(1)Explanations based on the patients' stage of\\nreadinessJudge what I (doctor) say\\nand how I (doctor) say it based on long-standing\\nrelationship with the patients5Provide\\ninformation based on the patient’s reaction and\\nacceptance of their medical condition5(2)Respect\\nfor the preferences of patients and their\\nfamiliesProvide easy-to-understand\\nexplanations without using technical terms5Provide explanations so that patients/families do not\\ndevelop misconceptions4Convey information\\ncarefully and clearly3Provide a written\\nexplanation2Divide bad news into\\ncomponents14Respect for the intentions of\\npatients and families(1)Respect for the\\nwill of patients and familiesRespect and support the patient’s will and desired\\ntreatment6If the patient’s\\ncondition is severe, respect the views of the family and\\nprimary caregiver2(2)Setting goals in accordance\\nwith the preferences of the patients and\\nfamiliesListen to the needs of\\npatients and their families and determine actionable\\ntreatment goals3Explore the intentions of\\npatients and their families through\\nconversations1If there is no\\ndifference between the advantages and disadvantages of a\\ntreatment choice, the choice is purely made by the\\npatient15Directing decision-making and considering\\nfairness(1)Considering the direction and\\nfairness of treatment options recommended by\\ndoctorsExplain treatment options\\nwith some direction while also providing options that\\nthe doctor finds beneficial8In some cases,\\noffer personal opinion as a doctor2Dare to treat\\nthe patient with paternalism2(2)Consideration for\\npatients to make impartial decisionsCommunicate with the knowledge that the\\ndoctor’s manner of speaking influences the patients’\\ndecisions5Explain the opinions of\\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\\nfamily(1)Explaining with consideration\\nfor emotionsTalk in a way that\\nallows the patient to have hope5Convey bad news\\naccording to the protocol1Consider that patients\\ndo not have to listen to explanations while always\\nlooking at bad information1Divulge bad news in\\nstages1(2)Showing a supportive attitude to the\\npatientApproach the patient with a\\nnon-abandoning attitude2Show that the team\\nsupports the patient1(3)Attitude when facing patients\\nand their familiesInteract with\\npatients on an equal footing2Talk while\\nlooking into the patient’s eyes1Interact with\\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\\ntreatmentsSupport fluctuations in\\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\nRole of hematologists in providing\\ndirect decision support (n =\\n11).\\n*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\nThis category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\nThis category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\nSeveral participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\nAs shown in Table 2,\\nwe identified 7 categories related to hematologists’ roles in providing direct\\ndecision support to patients: (1) preparing patients before informed consent,\\n(2) selecting the information to convey, (3) choosing a method for conveying\\nthis information, (4) respecting the intentions of patients and their families,\\n(5) directing decision-making and considering fairness, (6) considering the\\nemotional aspects of patients and their families, and (7) providing support\\nafter discussing treatment options.Table 2.Role of hematologists in providing\\ndirect decision support (n =\\n11).CategorySubcategoryCoden1Preparing patients before\\nIC*(1)Gathering preliminary\\ninformation for explaining issues to patientsObtain information from family members to\\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\\nfamily background1Talk to the family about\\nwhat you will tell the patient1(2)Encouraging family\\nmembers to participate in decision-makingSet up a place to provide information to\\npatients, including family members3Discuss\\nlife-prolonging treatments with the patient’s\\nfamily1(3)Creating an environment for concentrating on\\nICDiscuss in a private\\nroom2Take sufficient time1Leave the\\nPHS*2 with\\nthe nurse so that the discussion is not\\ndisturbed1(4)Providing advance information on the risk\\nof relapseWhen the patient reaches\\ncomplete remission, explain the risk of relapse and\\nprepare the patient for relapse12Selecting the\\ninformation to convey(1)Choice of\\ninformation to provide to patientsIn many cases, do not give the patient a\\nprognosis9Depending on the\\npatient’s medical condition, I (doctor) may not tell the\\ntruth5Modify the message according to the\\npatient’s life stage3(2)Providing patients and families\\nwith information to help them make treatment\\ndecisionsCommunicate the\\nadvantages and disadvantages of treatment\\noptions9Tell patients and their families the\\ninformation they need to make treatment\\ndecisions6Inform patients of all\\ntreatment options4Communicate not only the\\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\\nhelp them understand the patients' current\\nsituationInform family members\\nabout the possibility of sudden changes in the medical\\ncondition and specific prognosis9Provide\\npatients with information related to what they imagine\\nwill happen in the future5Provide patients with\\ninformation that will help them understand their current\\nmedical condition4Provide patients with a\\nvague idea of their prognosis3Communicate the\\nprognosis to patients using concrete numbers23Choosing\\nhow to communicate information(1)Explanations based on the patients' stage of\\nreadinessJudge what I (doctor) say\\nand how I (doctor) say it based on long-standing\\nrelationship with the patients5Provide\\ninformation based on the patient’s reaction and\\nacceptance of their medical condition5(2)Respect\\nfor the preferences of patients and their\\nfamiliesProvide easy-to-understand\\nexplanations without using technical terms5Provide explanations so that patients/families do not\\ndevelop misconceptions4Convey information\\ncarefully and clearly3Provide a written\\nexplanation2Divide bad news into\\ncomponents14Respect for the intentions of\\npatients and families(1)Respect for the\\nwill of patients and familiesRespect and support the patient’s will and desired\\ntreatment6If the patient’s\\ncondition is severe, respect the views of the family and\\nprimary caregiver2(2)Setting goals in accordance\\nwith the preferences of the patients and\\nfamiliesListen to the needs of\\npatients and their families and determine actionable\\ntreatment goals3Explore the intentions of\\npatients and their families through\\nconversations1If there is no\\ndifference between the advantages and disadvantages of a\\ntreatment choice, the choice is purely made by the\\npatient15Directing decision-making and considering\\nfairness(1)Considering the direction and\\nfairness of treatment options recommended by\\ndoctorsExplain treatment options\\nwith some direction while also providing options that\\nthe doctor finds beneficial8In some cases,\\noffer personal opinion as a doctor2Dare to treat\\nthe patient with paternalism2(2)Consideration for\\npatients to make impartial decisionsCommunicate with the knowledge that the\\ndoctor’s manner of speaking influences the patients’\\ndecisions5Explain the opinions of\\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\\nfamily(1)Explaining with consideration\\nfor emotionsTalk in a way that\\nallows the patient to have hope5Convey bad news\\naccording to the protocol1Consider that patients\\ndo not have to listen to explanations while always\\nlooking at bad information1Divulge bad news in\\nstages1(2)Showing a supportive attitude to the\\npatientApproach the patient with a\\nnon-abandoning attitude2Show that the team\\nsupports the patient1(3)Attitude when facing patients\\nand their familiesInteract with\\npatients on an equal footing2Talk while\\nlooking into the patient’s eyes1Interact with\\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\\ntreatmentsSupport fluctuations in\\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\nRole of hematologists in providing\\ndirect decision support (n =\\n11).\\n*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\nThis category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\nThis category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\nSeveral participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\n[SUBTITLE] Role of Hematologists in Multidisciplinary Collaboration and Providing\\nDecision Support [SUBSECTION] We identified the following 5 subcategories related to the second main study\\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\\ncommunicating with other professionals, (2) gathering information from them, (3)\\nproviding information to them, (4) managing the entire medical team, and (5)\\nencouraging nurses to actively participate with patients throughout the\\ndecision-making process (Table 3). More than half of the participants recognized the\\nhematologist’s role in communicating with other professionals and obtaining\\ninformation about the patient (eg, the patient’s background and how to help)\\nthroughout this process. Others said that they actively encourage nurses to\\nsupport the patients’ decisions and provide additional information during the\\ndiscussions or explain the prognoses and future treatment plans to other\\nprofessionals.Table\\n3.The role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\\nprofessionals, such as nursesMake opportunities to hold discussions with\\nmultiple professionals2Gathering\\ninformation from other professionalsObtain information from nurses about family\\nbackground and family supportObtain information about the patients’\\nthoughts and intentions from other\\nprofessionalsObtain\\ninformation about the patients’ social background from\\nother professionalsObtain\\ninformation about the patient’s ADL status from other\\nprofessionals3Providing\\ninformation to other professionalsExplain the patient’s prognosis and future treatment\\nplans to other professionalsExplain the concerns of the patient’s\\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\\nhematologists make comprehensive decisions based on\\ninformation from other professionalsCreate a support system for\\npatients5Encouraging nurses to actively\\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\\npatients’ decision-makingConfirm whether there is any additional information or\\nthoughts to convey to nurses during discussions about\\ntreatment decisions with the\\npatientADL\\n= Activities of Daily\\nLiving.\\nThe role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).\\nADL\\n= Activities of Daily\\nLiving.\\nWe identified the following 5 subcategories related to the second main study\\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\\ncommunicating with other professionals, (2) gathering information from them, (3)\\nproviding information to them, (4) managing the entire medical team, and (5)\\nencouraging nurses to actively participate with patients throughout the\\ndecision-making process (Table 3). More than half of the participants recognized the\\nhematologist’s role in communicating with other professionals and obtaining\\ninformation about the patient (eg, the patient’s background and how to help)\\nthroughout this process. Others said that they actively encourage nurses to\\nsupport the patients’ decisions and provide additional information during the\\ndiscussions or explain the prognoses and future treatment plans to other\\nprofessionals.Table\\n3.The role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\\nprofessionals, such as nursesMake opportunities to hold discussions with\\nmultiple professionals2Gathering\\ninformation from other professionalsObtain information from nurses about family\\nbackground and family supportObtain information about the patients’\\nthoughts and intentions from other\\nprofessionalsObtain\\ninformation about the patients’ social background from\\nother professionalsObtain\\ninformation about the patient’s ADL status from other\\nprofessionals3Providing\\ninformation to other professionalsExplain the patient’s prognosis and future treatment\\nplans to other professionalsExplain the concerns of the patient’s\\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\\nhematologists make comprehensive decisions based on\\ninformation from other professionalsCreate a support system for\\npatients5Encouraging nurses to actively\\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\\npatients’ decision-makingConfirm whether there is any additional information or\\nthoughts to convey to nurses during discussions about\\ntreatment decisions with the\\npatientADL\\n= Activities of Daily\\nLiving.\\nThe role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).\\nADL\\n= Activities of Daily\\nLiving.\\n[SUBTITLE] Expected Roles of Other Professionals in Providing Decision Support [SUBSECTION] We extracted the following 4 categories related to the roles that hematologists\\nexpected other professionals to fulfill: (1) sharing patient information and\\ncare routines, (2) sharing information from treatment discussions, (3) providing\\nemotional support to patients and their families, and (4) assembling a medical\\nteam that understands hematological malignancies (Table 4). Nearly half of the\\nparticipants indicated that they would like other health care providers to\\nconvey the patients’ thoughts and intentions that they knew to the doctors\\ninvolved. They also expected other professionals to share information regarding\\ntheir patients and the way in which they cared for their patients.Table 4.Expected\\nroles of other professionals in decision support (n =\\n11).SubcategoryCodenSharing patient information and care\\nroutinesListen to the patients’\\nthoughts and wishes and communicate them to the\\ndoctor5Suggest knowledge that doctors may not know\\nand explain how to care for patients in the right\\nway5Point out the problems patients have with treatment\\ndecisions2Sharing information from the\\ntreatment discussionAttend\\ntreatment discussions and understand the patients’\\nfeelings and situation together4Read records of the\\nprocess of informed consent and share the\\ninformation1Providing patients and their families with\\nemotional supportSupport the\\nemotions of patients and their families4Assembling\\na medical team that understands hematological\\nmalignancyWork to ensure that the\\nmedical team understands hematological\\nmalignancy1\\nExpected\\nroles of other professionals in decision support (n =\\n11).\\nWe extracted the following 4 categories related to the roles that hematologists\\nexpected other professionals to fulfill: (1) sharing patient information and\\ncare routines, (2) sharing information from treatment discussions, (3) providing\\nemotional support to patients and their families, and (4) assembling a medical\\nteam that understands hematological malignancies (Table 4). Nearly half of the\\nparticipants indicated that they would like other health care providers to\\nconvey the patients’ thoughts and intentions that they knew to the doctors\\ninvolved. They also expected other professionals to share information regarding\\ntheir patients and the way in which they cared for their patients.Table 4.Expected\\nroles of other professionals in decision support (n =\\n11).SubcategoryCodenSharing patient information and care\\nroutinesListen to the patients’\\nthoughts and wishes and communicate them to the\\ndoctor5Suggest knowledge that doctors may not know\\nand explain how to care for patients in the right\\nway5Point out the problems patients have with treatment\\ndecisions2Sharing information from the\\ntreatment discussionAttend\\ntreatment discussions and understand the patients’\\nfeelings and situation together4Read records of the\\nprocess of informed consent and share the\\ninformation1Providing patients and their families with\\nemotional supportSupport the\\nemotions of patients and their families4Assembling\\na medical team that understands hematological\\nmalignancyWork to ensure that the\\nmedical team understands hematological\\nmalignancy1\\nExpected\\nroles of other professionals in decision support (n =\\n11).\",\n \"We conducted semi-structured interviews involving 11 hematologists. The average\\nparticipant age was 44 years (range: 35-54 years), whereas the average\\nspecialization experience was 16 years (range: 7-26 years). Additional details\\nare listed in Table\\n1. The average length of the interviews was 27 minutes (range: 17-46\\nminutes).Table\\n1.Characteristics of doctors and interview\\ndetails.HematologistsTotaln\\n= 11Gender—Male10Female1Workplace—University hospital7Cancer center2General\\nhospital2Age (years)43.9 (34-54)*Experience in specialty (years)16.1\\n(7-26)*∗Mean(Range).\\nCharacteristics of doctors and interview\\ndetails.\\n∗Mean(Range).\",\n \"As shown in Table 2,\\nwe identified 7 categories related to hematologists’ roles in providing direct\\ndecision support to patients: (1) preparing patients before informed consent,\\n(2) selecting the information to convey, (3) choosing a method for conveying\\nthis information, (4) respecting the intentions of patients and their families,\\n(5) directing decision-making and considering fairness, (6) considering the\\nemotional aspects of patients and their families, and (7) providing support\\nafter discussing treatment options.Table 2.Role of hematologists in providing\\ndirect decision support (n =\\n11).CategorySubcategoryCoden1Preparing patients before\\nIC*(1)Gathering preliminary\\ninformation for explaining issues to patientsObtain information from family members to\\nunderstand the patient’s preferences in advance3Check if home care is possible1Understand the\\nfamily background1Talk to the family about\\nwhat you will tell the patient1(2)Encouraging family\\nmembers to participate in decision-makingSet up a place to provide information to\\npatients, including family members3Discuss\\nlife-prolonging treatments with the patient’s\\nfamily1(3)Creating an environment for concentrating on\\nICDiscuss in a private\\nroom2Take sufficient time1Leave the\\nPHS*2 with\\nthe nurse so that the discussion is not\\ndisturbed1(4)Providing advance information on the risk\\nof relapseWhen the patient reaches\\ncomplete remission, explain the risk of relapse and\\nprepare the patient for relapse12Selecting the\\ninformation to convey(1)Choice of\\ninformation to provide to patientsIn many cases, do not give the patient a\\nprognosis9Depending on the\\npatient’s medical condition, I (doctor) may not tell the\\ntruth5Modify the message according to the\\npatient’s life stage3(2)Providing patients and families\\nwith information to help them make treatment\\ndecisionsCommunicate the\\nadvantages and disadvantages of treatment\\noptions9Tell patients and their families the\\ninformation they need to make treatment\\ndecisions6Inform patients of all\\ntreatment options4Communicate not only the\\nbad news, but also ways of improvement3Also convey the option of non-treatment3(3)Providing patients and families with information to\\nhelp them understand the patients' current\\nsituationInform family members\\nabout the possibility of sudden changes in the medical\\ncondition and specific prognosis9Provide\\npatients with information related to what they imagine\\nwill happen in the future5Provide patients with\\ninformation that will help them understand their current\\nmedical condition4Provide patients with a\\nvague idea of their prognosis3Communicate the\\nprognosis to patients using concrete numbers23Choosing\\nhow to communicate information(1)Explanations based on the patients' stage of\\nreadinessJudge what I (doctor) say\\nand how I (doctor) say it based on long-standing\\nrelationship with the patients5Provide\\ninformation based on the patient’s reaction and\\nacceptance of their medical condition5(2)Respect\\nfor the preferences of patients and their\\nfamiliesProvide easy-to-understand\\nexplanations without using technical terms5Provide explanations so that patients/families do not\\ndevelop misconceptions4Convey information\\ncarefully and clearly3Provide a written\\nexplanation2Divide bad news into\\ncomponents14Respect for the intentions of\\npatients and families(1)Respect for the\\nwill of patients and familiesRespect and support the patient’s will and desired\\ntreatment6If the patient’s\\ncondition is severe, respect the views of the family and\\nprimary caregiver2(2)Setting goals in accordance\\nwith the preferences of the patients and\\nfamiliesListen to the needs of\\npatients and their families and determine actionable\\ntreatment goals3Explore the intentions of\\npatients and their families through\\nconversations1If there is no\\ndifference between the advantages and disadvantages of a\\ntreatment choice, the choice is purely made by the\\npatient15Directing decision-making and considering\\nfairness(1)Considering the direction and\\nfairness of treatment options recommended by\\ndoctorsExplain treatment options\\nwith some direction while also providing options that\\nthe doctor finds beneficial8In some cases,\\noffer personal opinion as a doctor2Dare to treat\\nthe patient with paternalism2(2)Consideration for\\npatients to make impartial decisionsCommunicate with the knowledge that the\\ndoctor’s manner of speaking influences the patients’\\ndecisions5Explain the opinions of\\nother doctors and the second opinion system16Considering the emotional aspects of the patient and\\nfamily(1)Explaining with consideration\\nfor emotionsTalk in a way that\\nallows the patient to have hope5Convey bad news\\naccording to the protocol1Consider that patients\\ndo not have to listen to explanations while always\\nlooking at bad information1Divulge bad news in\\nstages1(2)Showing a supportive attitude to the\\npatientApproach the patient with a\\nnon-abandoning attitude2Show that the team\\nsupports the patient1(3)Attitude when facing patients\\nand their familiesInteract with\\npatients on an equal footing2Talk while\\nlooking into the patient’s eyes1Interact with\\nthe patient’s family on an equal footing17Supporting after discussing treatments(1)Emotional support after discussing\\ntreatmentsSupport fluctuations in\\nthe patient’s mind after discussing treatments2Call out the family and assess their reactions1*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\nRole of hematologists in providing\\ndirect decision support (n =\\n11).\\n*1IC\\n= informed consent; *2PHS = Personal Handyphone\\nSystem (A phone used by doctors to communicate with staff in the\\nhospital).\\n[SUBTITLE] Preparing Patients Before Informed Consent (IC) [SUBSECTION] This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\nThis category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\\n[SUBTITLE] Selecting the Information To Convey [SUBSECTION] This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\nThis category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\\n[SUBTITLE] Choosing the Way In Which To Communicate Information [SUBSECTION] This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\nThis category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\\n[SUBTITLE] Respect for The Intentions Of Patients And Families [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\nThis category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\\n[SUBTITLE] Directing decision-making and considering fairness [SUBSECTION] This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\nThis category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\\n[SUBTITLE] Considering the Emotional Aspects Of The Patient And Family [SUBSECTION] This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\nThis category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\\n[SUBTITLE] Providing support after discussing treatments [SUBSECTION] Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\\nSeveral participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\",\n \"This category comprised 4 subcategories. Several participants talked about\\npreparing patients and families for IC discussions, including “Gathering\\npreliminary information for explaining issues to patients” and “Encouraging\\nfamily members to participate in decision-making.” As such, physicians were\\nexpected to prepare patients before IC discussions. For instance, they would\\nprovide patients with information regarding the risk of relapse.\",\n \"This category comprised 3 subcategories. In subcategory (1) “Choice of\\ninformation to provide to patients,” more than half of the participants said\\nthat they did not provide their patients with prognoses. More importantly,\\nmore than half said they may not even tell their patients the truth,\\ndepending on the severity of the illness, to avoid a loss of hope; they were\\nvery selective about the information they chose to provide to their\\npatients.\\nIn subcategory (2) “Providing patients and families with information to help\\nthem make treatment decisions,” more than half of the participants\\nidentified “Communicating the advantages and disadvantages of treatment\\noptions” and “Telling patients and their families the information they need\\nto make treatment decisions.” They provided information they believed would\\nhelp patients and their families fully understand all the options before\\nmaking a final decision.\\nIn subcategory (3), “Providing patients and families with information to help\\nthem understand the patients’ current situation,” a large number said they\\nwould inform family members about the possibility of sudden changes and\\nspecific prognoses, with more than half saying they would explain the issues\\nso that patients could visualize what may happen to them in the future.\\nGenerally, participants provided information to help patients and their\\nfamilies clearly understand their current situations.\",\n \"This category comprised 2 subcategories. Nearly half of participants\\nexpressed the ideas of “I determine what I say and how I say it based on our\\nlong-standing relationship with the patients” and “Providing information\\nbased on the patient’s reaction and acceptance of their medical condition.”\\nThey carefully considered what they discussed and how they discussed it,\\nespecially considering their long-standing association with their patients.\\nIn this regard, they conveyed information based on how they thought patients\\nwould react and their level of acceptance. They also provided explanations\\naccording to what they believed patients intended and thought depending on\\ntheir respective stages of readiness. Moreover, nearly half said that they\\nwould explain issues without using jargon, thereby ensuring that the\\ninformation provided was easy to understand. Essentially, participants\\nremained cognizant of the need to communicate both correctly and\\nappropriately with patients and their families.\",\n \"This category comprised 2 subcategories. In subcategory (1) “Respect for the\\nwill of patients and families,” more than half of the participants said they\\nhelped patients and families make decisions by supporting and respecting\\ntheir wishes and desired treatments. In subcategory (2) “Setting goals in\\naccordance with the preferences of the patients and families,” several\\nparticipants said that they helped patients and families set feasible goals\\nafter listening to their wishes.\",\n \"This category comprised 2 subcategories. In subcategory (1) “Considering the\\ndirection and fairness of treatment options recommended by doctors,” more\\nthan half of the participants said they explain treatment options with some\\ndirection while also providing options they find most beneficial. They\\ntalked about giving weight to treatment options and facilitating the\\ndecision-making process. Furthermore, they attempted to present information\\nin a fair manner, being aware that the way they talk can affect the way\\npatients perceive their treatment, as shown in subcategory (2).\",\n \"This category comprised 3 subcategories. In subcategory (1)“explaining with\\nconsideration for emotions,” nearly half of the participants said they would\\nattempt to instill hope among patients, and hence, provide explanations\\nconsidering the related emotional aspects. In subcategory (2) “showing a\\nsupportive attitude to the patient”, several participants also expressed\\nadopting a supportive stance to reassure patients that they were not\\nabandoning them or their families. In subcategory (3) “attitude when facing\\npatients and their families”, several participants talked about the doctor’s\\nattitude toward patients, such as speaking with patients and families as\\nequals and looking them in the eye.\",\n \"Several participants said they informed patients about bad news and future\\ntreatments, after which they provided emotional support in consideration of\\nany emotional variabilities.\",\n \"We identified the following 5 subcategories related to the second main study\\ntopic—the roles of hematologists in multidisciplinary collaboration: (1)\\ncommunicating with other professionals, (2) gathering information from them, (3)\\nproviding information to them, (4) managing the entire medical team, and (5)\\nencouraging nurses to actively participate with patients throughout the\\ndecision-making process (Table 3). More than half of the participants recognized the\\nhematologist’s role in communicating with other professionals and obtaining\\ninformation about the patient (eg, the patient’s background and how to help)\\nthroughout this process. Others said that they actively encourage nurses to\\nsupport the patients’ decisions and provide additional information during the\\ndiscussions or explain the prognoses and future treatment plans to other\\nprofessionals.Table\\n3.The role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).SubcategoryCode1Communicating with other professionalsValue good communication with other\\nprofessionals, such as nursesMake opportunities to hold discussions with\\nmultiple professionals2Gathering\\ninformation from other professionalsObtain information from nurses about family\\nbackground and family supportObtain information about the patients’\\nthoughts and intentions from other\\nprofessionalsObtain\\ninformation about the patients’ social background from\\nother professionalsObtain\\ninformation about the patient’s ADL status from other\\nprofessionals3Providing\\ninformation to other professionalsExplain the patient’s prognosis and future treatment\\nplans to other professionalsExplain the concerns of the patient’s\\nfamily to other professionals4Managing the entire medical teamManage the medical team, in which\\nhematologists make comprehensive decisions based on\\ninformation from other professionalsCreate a support system for\\npatients5Encouraging nurses to actively\\nparticipate in the patient’s decision-makingEncourage nurses to participate in the\\npatients’ decision-makingConfirm whether there is any additional information or\\nthoughts to convey to nurses during discussions about\\ntreatment decisions with the\\npatientADL\\n= Activities of Daily\\nLiving.\\nThe role of hematologists in multidisciplinary\\ncollaboration and decision support (n =\\n11).\\nADL\\n= Activities of Daily\\nLiving.\",\n \"We extracted the following 4 categories related to the roles that hematologists\\nexpected other professionals to fulfill: (1) sharing patient information and\\ncare routines, (2) sharing information from treatment discussions, (3) providing\\nemotional support to patients and their families, and (4) assembling a medical\\nteam that understands hematological malignancies (Table 4). Nearly half of the\\nparticipants indicated that they would like other health care providers to\\nconvey the patients’ thoughts and intentions that they knew to the doctors\\ninvolved. They also expected other professionals to share information regarding\\ntheir patients and the way in which they cared for their patients.Table 4.Expected\\nroles of other professionals in decision support (n =\\n11).SubcategoryCodenSharing patient information and care\\nroutinesListen to the patients’\\nthoughts and wishes and communicate them to the\\ndoctor5Suggest knowledge that doctors may not know\\nand explain how to care for patients in the right\\nway5Point out the problems patients have with treatment\\ndecisions2Sharing information from the\\ntreatment discussionAttend\\ntreatment discussions and understand the patients’\\nfeelings and situation together4Read records of the\\nprocess of informed consent and share the\\ninformation1Providing patients and their families with\\nemotional supportSupport the\\nemotions of patients and their families4Assembling\\na medical team that understands hematological\\nmalignancyWork to ensure that the\\nmedical team understands hematological\\nmalignancy1\\nExpected\\nroles of other professionals in decision support (n =\\n11).\",\n \"In this study, hematologists were interviewed to clarify their role and expectations\\nof other professionals. The categories of hematologists’ direct role in\\ndecision-making were extracted, including preparation and consideration of the\\nsituation in which information about decision-making is communicated and emotional\\nsupport after the information is communicated.\\nHematologists presented treatment options based on a variety of factors, including\\nthe patient’s goals, emotions, the preparedness for the disease, and the severity of\\nthe disease. During this process, the hematologists worked with other professionals\\nto understand specific information about the patient and family, including their\\nbackground, goals, and intentions. They also said that they would decide what\\ninformation to convey and carefully explain the treatment and medical condition,\\nwithout using jargon, so that the patient and family could easily understand. ASCO\\nguidelines recommend that physicians provide information that takes into account\\nindividual patient interests and preferences.7 Although the present results\\nare data obtained before these guidelines were developed, the hematologists who\\nparticipated in this study were taking the actions recommended in the guidelines.\\nThe present results reveal more specific behaviors of hematologists with regard to\\nwhat is recommended in the guidelines. For patients facing complex and risky\\ntreatment choices, it is important for hematologist to select information according\\nto the patient’s wishes and to convey information in easy-to-understand terms.\\nHematologists were expected other professionalsto pay attention to the emotional\\naspects of the patient and family to minimize their anxiety. In addition, they\\nexpected nurses and other professionals to listen to the patient and family’s wishes\\nand thoughts about treatment, share them with the physician, and serve as emotional\\nsupport for the patient and family. Moreover, hematologists were expected to\\nunderstand the patient’s wishes and preferences, explain complex and\\ndifficult-to-understand treatments and medical conditions, and present options that\\nwere consistent with the patient’s wishes. ASCO guidelines also emphasize the need\\nto form a rapport with patients and their families and to assess and support coping\\nneeds.7 The hematologists who participated in this study wanted to\\nimplement this in collaboration with other professionals. The wishes and intentions\\nof patients and their families change as the disease progresses or symptoms worsen.\\nTherefore, it is thought that by having physicians and other professions work\\ntogether to understand the wishes of patients/families at that time, decision-making\\nsupport in accordance with the wishes of patients/families will become possible.\\nLeukemia and lymphoma may be treated through chemotherapy, even in cases of relapse\\nand during the refractory phase. This makes it difficult to determine when to\\ntransition from curative treatments to those more focused on quality of life\\n(QOL).13,14 Previous studies have reported that hematologists are reluctant\\nto discuss EOL issues while primary disease treatments are still possible.15,16 A number of\\nthe hematologists who participated in the study had a variety of ways of explaining\\nthe prognosis to patients and their families. For example, in some cases, they\\ninformed the family about the possibility of a sudden change in prognosis, including\\nspecific figures, while in other cases, depending on the severity of the disease and\\nthe stage of life, they did not inform them of the prognosis at all or withheld the\\nfact. Furthermore, more than half of the respondents stated that they would explain\\nthe treatment that they thought would be most beneficial to the patient.\\nA number of the hematologists who participated in this study also had a long history\\nof working with patients to understand their individual backgrounds, and what they\\ntalked about and how they communicate, while also having a direction that the\\nhematologists thought was suitable, and gave the patients some direction as they\\nspoke. This suggested that hematologists attempt to tailor their decision-making\\nsupport to the needs of their patients. On the other hand, hematologists infer\\npatients' needs based on their previous experience, which runs the risk of making\\nsuggestions that are not in line with the patients’ original needs. Previous studies\\nhave reported that hematologic oncology patients take a passive role in the\\ntreatment decision-making process due to the complexity of their treatment and\\ndisease17 and expect their physicians to take a paternalistic\\nrole.16 When hematologists provide some treatment direction and\\npaternalistic response to hematologic oncology patients with these characteristics,\\npatients may not fully understand the advantages and disadvantages of treatment and\\nmay not make an informed choice. If treatment is responded to, the patient’s quality\\nof life improves, but if treatment is not effective, the patient may die in the\\nhospital during active treatment. In addition, patients may not be able to live the\\nlife they want, with increased risk of infection and prolonged\\nhospitalization.1 Hematologists need to have frank discussions with patients,\\nkeeping in mind the direction they think is best for the patient, and working with\\nother professionals to understand the patient’s wishes.\\nHematologists who participated in this study indicated that the role they expected of\\nmultidisciplinary professionals was to inform them about patient thoughts,\\ninformation, and methods of care that the physicians were unaware of. They also\\nexpected other professions to play a role in providing emotional support to patients\\nand their families. In a previous study of oncologists, including hematologists,\\nlack of information about the patient was the most frequently cited factor hindering\\npatient involvement in decision-making.18 Physicians also recognized\\nthat having a third party present during decision-making discussions supported\\npatient decision-making and facilitated involvement and reflection on treatment\\ndecisions.18 In that previous study, physicians emphasized the\\nimportance of obtaining information about the patient and the presence of a third\\nparty in the decision-making situation, which was similar to the findings of the\\npresent study. This confirms the importance of a multidisciplinary approach to\\nfacilitate decision-making in accordance with the patient’s wishes. In making\\ntreatment decisions, it is necessary to organize information not only from a medical\\nperspective, but also based on the patient’s priorities and values. Obtaining such\\ninformation is difficult for hematologists alone, and multidisciplinary\\ncollaboration is important. By understanding the patient’s needs and organizing the\\ninformation necessary for decision-making, it is expected that the discussion\\nbetween the attending hematologist and the patient will be deepened.9\\nFinally, it is important to note that the survey for this study was conducted in 2011\\nand there was not much collaboration with the palliative care team,15 the\\nparticipants in this study did not mention the role they expected to play regarding\\npalliative care physicians. Recently, however, the effects of early palliative care\\nhave been reported in clinical hematologic oncology. A study evaluating the effects\\nof early introduction of palliative care and continuous psychological support by a\\nmultidisciplinary team in patients with acute leukemia reported a reduction in\\nsymptoms due to acute stress reactions compared to the usual care group.19 It has also\\nbeen reported that patients undergoing hematopoietic stem cell transplantation who\\nreceived regular twice-weekly visits by a palliative care physician had a higher QOL\\nat 3 months of hospitalization compared to the group that received usual\\ncare.20 In patients with leukemia and lymphoma, where the potential for\\nrapid change is high, it is expected that hematologists and palliative care\\nspecialists will support patients and begin discussing long-term goals early, in\\ncase treatment goals suddenly change.21\\nThis study had some limitations. The first is that the data collection and analysis\\nwere conducted between 2011 and 2012. Therefore, because treatment outcomes for\\nhematologic tumors have improved compared to a decade ago, the data cannot be\\ndirectly applied to decision support for patients with such tumors and their\\nfamilies today. However, the patients with treatment-resistant hematologic tumors\\nhave not changed significantly in their characteristics or treatment\\nenvironment,1,22 and it is significant to clarify the hematologist’s role in\\ndecision support. Second, the use of the opportunistic sampling makes it difficult\\nto generalize the findings. In this study, we collected the opinions of\\nhematologists at hospitals where hematologic oncology patients are mainly treated,\\nsuch as cancer hospitals and university hospitals in Japan, and we believe that we\\nwere able to collect diverse opinions regarding the role of decision support as\\nperceived by hematologists. Based on the present study, research is needed to\\nclarify the decision support roles of other professionals who care for patients with\\nleukemia and lymphoma, including nurses. This would make it possible to construct a\\nmodel of shared decision-making support for hematologic oncology patients through\\nmultidisciplinary collaboration.\",\n \"Click here for additional data file.\\nSupplemental Material for Qualitative Analysis of the Roles of Physicians and\\nNurses in Providing Decision Support to Patients With Relapsed or Refractory\\nLeukemia and Lymphoma by Miharu Morikawa, and Yuki Shirai in Cancer Control\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["intro","methods","results",null,null,null,null,null,null,null,null,null,null,null,"discussion","supplementary-material"],"string":"[\n \"intro\",\n \"methods\",\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["hematologic diseases","shared decision-making","patient care team","physician’s role","communication"],"string":"[\n \"hematologic diseases\",\n \"shared decision-making\",\n \"patient care team\",\n \"physician’s role\",\n \"communication\"\n]"}}},{"rowIdx":380,"cells":{"title":{"kind":"string","value":"Defective Desmosomal Adhesion Causes Arrhythmogenic Cardiomyopathy by Involving an Integrin-αVβ6/TGF-β Signaling Cascade."},"pmid":{"kind":"string","value":"36268721"},"background_abstract":{"kind":"string","value":"Arrhythmogenic cardiomyopathy (ACM) is characterized by progressive loss of cardiomyocytes with fibrofatty tissue replacement, systolic dysfunction, and life-threatening arrhythmias. A substantial proportion of ACM is caused by mutations in genes of the desmosomal cell-cell adhesion complex, but the underlying mechanisms are not well understood. In the current study, we investigated the relevance of defective desmosomal adhesion for ACM development and progression."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We mutated the binding site of DSG2 (desmoglein-2), a crucial desmosomal adhesion molecule in cardiomyocytes. This DSG2-W2A mutation abrogates the tryptophan swap, a central interaction mechanism of DSG2 on the basis of structural data. Impaired adhesive function of DSG2-W2A was confirmed by cell-cell dissociation assays and force spectroscopy measurements by atomic force microscopy. The DSG2-W2A knock-in mouse model was analyzed by echocardiography, ECG, and histologic and biomolecular techniques including RNA sequencing and transmission electron and superresolution microscopy. The results were compared with ACM patient samples, and their relevance was confirmed in vivo and in cardiac slice cultures by inhibitor studies applying the small molecule EMD527040 or an inhibitory integrin-αVβ6 antibody."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The DSG2-W2A mutation impaired binding on molecular level and compromised intercellular adhesive function. Mice bearing this mutation develop a severe cardiac phenotype recalling the characteristics of ACM, including cardiac fibrosis, impaired systolic function, and arrhythmia. A comparison of the transcriptome of mutant mice with ACM patient data suggested deregulated integrin-αVβ6 and subsequent transforming growth factor-β signaling as driver of cardiac fibrosis. Blocking integrin-αVβ6 led to reduced expression of profibrotic markers and reduced fibrosis formation in mutant animals in vivo."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"We show that disruption of desmosomal adhesion is sufficient to induce a phenotype that fulfils the clinical criteria to establish the diagnosis of ACM, confirming the dysfunctional adhesion hypothesis. Deregulation of integrin-αVβ6 and transforming growth factor-β signaling was identified as a central step toward fibrosis. A pilot in vivo drug test revealed this pathway as a promising target to ameliorate fibrosis. This highlights the value of this model to discern mechanisms of cardiac fibrosis and to identify and test novel treatment options for ACM."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Mice","Animals","Cardiomyopathies","Arrhythmias, Cardiac","Integrins","Myocytes, Cardiac","Fibrosis","Transforming Growth Factor beta","Transforming Growth Factors","Arrhythmogenic Right Ventricular Dysplasia"],"string":"[\n \"Mice\",\n \"Animals\",\n \"Cardiomyopathies\",\n \"Arrhythmias, Cardiac\",\n \"Integrins\",\n \"Myocytes, Cardiac\",\n \"Fibrosis\",\n \"Transforming Growth Factor beta\",\n \"Transforming Growth Factors\",\n \"Arrhythmogenic Right Ventricular Dysplasia\"\n]"},"pmcid":{"kind":"string","value":"9674449"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"Detailed methods are available in the Expanded Methods in the Supplemental Material.\n[SUBTITLE] Human Heart Samples [SUBSECTION] Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\nHeart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\n[SUBTITLE] Animal Experiments [SUBSECTION] Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\nMouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\n[SUBTITLE] Statistics and Data Compilation [SUBSECTION] Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\nFigures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\n[SUBTITLE] Data Availability [SUBSECTION] RNA sequencing data were deposited in the public database GEO (\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.\nRNA sequencing data were deposited in the public database GEO (\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo [SUBSECTION] To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\nTo study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\n[SUBTITLE] DSG2-W2A Mutant Mice Resemble the Phenotype of ACM [SUBSECTION] Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\nGiven the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\n[SUBTITLE] Integrin-β6 Expression Is Altered in ACM Patient and DSG2-W2A Mouse Samples [SUBSECTION] Next, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\nNext, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\n[SUBTITLE] DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs [SUBSECTION] This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\nThis result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\n[SUBTITLE] Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts [SUBSECTION] In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\nIn contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\n[SUBTITLE] TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts [SUBSECTION] ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\nITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\n[SUBTITLE] Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β [SUBSECTION] On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\nOn the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["What Is New?","What Are the Clinical Implications?","Human Heart Samples","Animal Experiments","Statistics and Data Compilation","W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo","DSG2-W2A Mutant Mice Resemble the Phenotype of ACM","DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs","Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts","TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts","Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β","Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype","DSG2-W2A Mice as Model to Study ACM","DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement","Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM","Article Information","Acknowledgments","Sources of Funding","Supplemental Material"],"string":"[\n \"What Is New?\",\n \"What Are the Clinical Implications?\",\n \"Human Heart Samples\",\n \"Animal Experiments\",\n \"Statistics and Data Compilation\",\n \"W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo\",\n \"DSG2-W2A Mutant Mice Resemble the Phenotype of ACM\",\n \"DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs\",\n \"Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts\",\n \"TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts\",\n \"Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β\",\n \"Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype\",\n \"DSG2-W2A Mice as Model to Study ACM\",\n \"DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement\",\n \"Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM\",\n \"Article Information\",\n \"Acknowledgments\",\n \"Sources of Funding\",\n \"Supplemental Material\"\n]"},"other_sections_texts":{"kind":"list like","value":["The tryptophan residue at position 2 of the desmosomal adhesion molecule DSG2 (desmoglein-2) is central for its binding function and cell–cell adhesion in vitro and in vivo.\nMice with abrogated DSG2 binding function (DSG2-W2A) develop a cardiac phenotype recalling arrhythmogenic cardiomyopathy with fibrosis, impaired systolic function, ECG abnormalities, and ventricular arrhythmia.\nIncreased integrin-αVβ6–dependent transforming growth factor–β signaling was identified as a driver of fibrosis.","We provide a new mouse model reproducing major features of arrhythmogenic cardiomyopathy that can be applied to study disease mechanisms and test therapeutic approaches.\nThe model suggests that loss of mechanical cardiomyocyte coupling is a central and early event causing arrhythmogenic cardiomyopathy.\nOur pilot in vivo study highlights the applicability and potential to target integrin-αVβ6–dependent transforming growth factor–β release as new therapeutic approach to diminish the development of fibrosis in arrhythmogenic cardiomyopathy.","Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.","Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.","Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.","To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.","Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.","This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.","In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.","ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.","On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.","Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.","Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24","DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35","The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.","[SUBTITLE] Acknowledgments [SUBSECTION] The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\nThe authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\n[SUBTITLE] Sources of Funding [SUBSECTION] This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\nThis project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\n[SUBTITLE] Disclosures [SUBSECTION] Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\nDr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\n[SUBTITLE] Supplemental Material [SUBSECTION] Expanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48\nExpanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48","The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.","This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).","Expanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48"],"string":"[\n \"The tryptophan residue at position 2 of the desmosomal adhesion molecule DSG2 (desmoglein-2) is central for its binding function and cell–cell adhesion in vitro and in vivo.\\nMice with abrogated DSG2 binding function (DSG2-W2A) develop a cardiac phenotype recalling arrhythmogenic cardiomyopathy with fibrosis, impaired systolic function, ECG abnormalities, and ventricular arrhythmia.\\nIncreased integrin-αVβ6–dependent transforming growth factor–β signaling was identified as a driver of fibrosis.\",\n \"We provide a new mouse model reproducing major features of arrhythmogenic cardiomyopathy that can be applied to study disease mechanisms and test therapeutic approaches.\\nThe model suggests that loss of mechanical cardiomyocyte coupling is a central and early event causing arrhythmogenic cardiomyopathy.\\nOur pilot in vivo study highlights the applicability and potential to target integrin-αVβ6–dependent transforming growth factor–β release as new therapeutic approach to diminish the development of fibrosis in arrhythmogenic cardiomyopathy.\",\n \"Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\",\n \"Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\",\n \"Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\",\n \"To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\",\n \"Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\",\n \"This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\",\n \"In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\",\n \"ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\",\n \"On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\",\n \"Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\",\n \"Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\",\n \"DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\",\n \"The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.\",\n \"[SUBTITLE] Acknowledgments [SUBSECTION] The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\\nThe authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\\n[SUBTITLE] Sources of Funding [SUBSECTION] This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\\nThis project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\\n[SUBTITLE] Disclosures [SUBSECTION] Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\\nDr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\\n[SUBTITLE] Supplemental Material [SUBSECTION] Expanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\\nExpanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\",\n \"The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\",\n \"This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\",\n \"Expanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["What Is New?","What Are the Clinical Implications?","Methods","Human Heart Samples","Animal Experiments","Statistics and Data Compilation","Data Availability","Results","W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo","DSG2-W2A Mutant Mice Resemble the Phenotype of ACM","Integrin-β6 Expression Is Altered in ACM Patient and DSG2-W2A Mouse Samples","DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs","Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts","TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts","Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β","Discussion","Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype","DSG2-W2A Mice as Model to Study ACM","DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement","Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM","Article Information","Acknowledgments","Sources of Funding","Disclosures","Supplemental Material","Supplementary Material"],"string":"[\n \"What Is New?\",\n \"What Are the Clinical Implications?\",\n \"Methods\",\n \"Human Heart Samples\",\n \"Animal Experiments\",\n \"Statistics and Data Compilation\",\n \"Data Availability\",\n \"Results\",\n \"W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo\",\n \"DSG2-W2A Mutant Mice Resemble the Phenotype of ACM\",\n \"Integrin-β6 Expression Is Altered in ACM Patient and DSG2-W2A Mouse Samples\",\n \"DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs\",\n \"Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts\",\n \"TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts\",\n \"Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β\",\n \"Discussion\",\n \"Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype\",\n \"DSG2-W2A Mice as Model to Study ACM\",\n \"DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement\",\n \"Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM\",\n \"Article Information\",\n \"Acknowledgments\",\n \"Sources of Funding\",\n \"Disclosures\",\n \"Supplemental Material\",\n \"Supplementary Material\"\n]"},"all_sections_texts":{"kind":"list like","value":["The tryptophan residue at position 2 of the desmosomal adhesion molecule DSG2 (desmoglein-2) is central for its binding function and cell–cell adhesion in vitro and in vivo.\nMice with abrogated DSG2 binding function (DSG2-W2A) develop a cardiac phenotype recalling arrhythmogenic cardiomyopathy with fibrosis, impaired systolic function, ECG abnormalities, and ventricular arrhythmia.\nIncreased integrin-αVβ6–dependent transforming growth factor–β signaling was identified as a driver of fibrosis.","We provide a new mouse model reproducing major features of arrhythmogenic cardiomyopathy that can be applied to study disease mechanisms and test therapeutic approaches.\nThe model suggests that loss of mechanical cardiomyocyte coupling is a central and early event causing arrhythmogenic cardiomyopathy.\nOur pilot in vivo study highlights the applicability and potential to target integrin-αVβ6–dependent transforming growth factor–β release as new therapeutic approach to diminish the development of fibrosis in arrhythmogenic cardiomyopathy.","Detailed methods are available in the Expanded Methods in the Supplemental Material.\n[SUBTITLE] Human Heart Samples [SUBSECTION] Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\nHeart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\n[SUBTITLE] Animal Experiments [SUBSECTION] Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\nMouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\n[SUBTITLE] Statistics and Data Compilation [SUBSECTION] Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\nFigures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\n[SUBTITLE] Data Availability [SUBSECTION] RNA sequencing data were deposited in the public database GEO (\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.\nRNA sequencing data were deposited in the public database GEO (\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.","Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.","Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.","Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.","RNA sequencing data were deposited in the public database GEO (\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.","[SUBTITLE] W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo [SUBSECTION] To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\nTo study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\n[SUBTITLE] DSG2-W2A Mutant Mice Resemble the Phenotype of ACM [SUBSECTION] Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\nGiven the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\n[SUBTITLE] Integrin-β6 Expression Is Altered in ACM Patient and DSG2-W2A Mouse Samples [SUBSECTION] Next, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\nNext, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\n[SUBTITLE] DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs [SUBSECTION] This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\nThis result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\n[SUBTITLE] Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts [SUBSECTION] In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\nIn contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\n[SUBTITLE] TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts [SUBSECTION] ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\nITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\n[SUBTITLE] Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β [SUBSECTION] On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\nOn the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.","To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.","Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.","Next, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.","This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.","In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.","ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.","On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.","In this study, we generated a knock-in mouse model with defective binding function of the adhesion molecule DSG2, which demonstrated that impaired desmosomal adhesion is sufficient to induce a phenotype mimicking the characteristics of ACM. Our data suggest a cascade of defective desmosomal adhesion, disrupted ICD structure, and subsequent activation of ITGAV/B6 with profibrotic TGF-β signaling as important underlying mechanisms leading to this phenotype (Figure 7). Moreover, our pilot study indicates a beneficial effect of ITGAV/B6 inhibition by EMD treatment with regard to fibrosis and several ECG parameters, suggesting that this pathway can be targeted successfully by drug treatment.\nSchematic conclusion of data. DSG2 (desmoglein-2) –W2A mutation with loss of desmosomal adhesion leads to impaired intercalated disc (ICD) structure with deregulation of integrin-β6 (ITGB6) and enhanced heterodimerization with integrin-αV (ITGAV). The dimer efficiently binds to the extracellular matrix and activates transforming growth factor–β (TGF-β) by removal of the latency-associated peptide (LAP). Active TGF-β can then induce profibrotic downstream signaling by means of SMAD molecules. In our experiments, this cascade was blocked by different approaches to inhibit ITGAV/B6.\n[SUBTITLE] Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype [SUBSECTION] Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\nBecause of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\n[SUBTITLE] DSG2-W2A Mice as Model to Study ACM [SUBSECTION] Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\nOur results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\n[SUBTITLE] DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement [SUBSECTION] DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\nDSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\n[SUBTITLE] Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM [SUBSECTION] The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.\nThe ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.","Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.","Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24","DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35","The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.","[SUBTITLE] Acknowledgments [SUBSECTION] The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\nThe authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\n[SUBTITLE] Sources of Funding [SUBSECTION] This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\nThis project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\n[SUBTITLE] Disclosures [SUBSECTION] Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\nDr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\n[SUBTITLE] Supplemental Material [SUBSECTION] Expanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48\nExpanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48","The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.","This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).","Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.","Expanded Methods\nTables S1–S3\nFigures S1–S6\nReferences 42–48",""],"string":"[\n \"The tryptophan residue at position 2 of the desmosomal adhesion molecule DSG2 (desmoglein-2) is central for its binding function and cell–cell adhesion in vitro and in vivo.\\nMice with abrogated DSG2 binding function (DSG2-W2A) develop a cardiac phenotype recalling arrhythmogenic cardiomyopathy with fibrosis, impaired systolic function, ECG abnormalities, and ventricular arrhythmia.\\nIncreased integrin-αVβ6–dependent transforming growth factor–β signaling was identified as a driver of fibrosis.\",\n \"We provide a new mouse model reproducing major features of arrhythmogenic cardiomyopathy that can be applied to study disease mechanisms and test therapeutic approaches.\\nThe model suggests that loss of mechanical cardiomyocyte coupling is a central and early event causing arrhythmogenic cardiomyopathy.\\nOur pilot in vivo study highlights the applicability and potential to target integrin-αVβ6–dependent transforming growth factor–β release as new therapeutic approach to diminish the development of fibrosis in arrhythmogenic cardiomyopathy.\",\n \"Detailed methods are available in the Expanded Methods in the Supplemental Material.\\n[SUBTITLE] Human Heart Samples [SUBSECTION] Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\\nHeart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\\n[SUBTITLE] Animal Experiments [SUBSECTION] Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\\nMouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\\n[SUBTITLE] Statistics and Data Compilation [SUBSECTION] Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\\nFigures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\\n[SUBTITLE] Data Availability [SUBSECTION] RNA sequencing data were deposited in the public database GEO (\\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.\\nRNA sequencing data were deposited in the public database GEO (\\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.\",\n \"Heart samples of patients with ACM and healthy controls were derived from forensic autopsy. This study was conducted according to the tenets of the Declaration of Helsinki and approved by the local ethics committees (license number 494-16 from LMU Munich and license number 152/15 from Goethe University). Samples were fixed and embedded in paraffin according to standard procedures.\",\n \"Mouse experiments were approved by the Cantonal Veterinary Office of Basel-Stadt (license numbers 2973_32878 and 3070_32419). Mice were housed under specific pathogen-free conditions with standard chow and bedding with a 12-hour day/night cycle according to institutional guidelines. Animals of both sexes were applied without bias. For inhibitor treatments, mice from the same litter and sex were paired and allocated randomly to the groups.\",\n \"Figures were compiled with Adobe Photoshop CC 2017 and Adobe Illustrator CC 2017. Statistical computations were performed with Prism 8 (GraphPad Software). For comparison of 2 or multiple groups, distribution of data was analyzed by a Shapiro-Wilk normality test, and group variances were analyzed by an F-test or a Brown-Forsythe test, respectively. According to the results of these tests, a parametric or nonparametric test with or without Welch correction for unequal variances was applied. For Kaplan-Meier survival analysis, significance was assessed by the Gehan-Breslow-Wilcoxon test. The statistical test used to compare the respective data sets is described in the corresponding figure legend. Statistical significance was assumed at P<0.05. Unless otherwise stated, data are presented as dot blot, with each dot representing the mean of the respective technical replicates of 1 biological replicate. Each animal or independent seeding of cells was taken as biological replicate, or as indicated in the legend. The mean value of these dots is shown as bar diagram ± SD.\",\n \"RNA sequencing data were deposited in the public database GEO (\\naccession number GSE181868). Additional data pertaining to the current article are available from the corresponding author upon request.\",\n \"[SUBTITLE] W2A Mutation Abrogates DSG2 Interactions and Impairs Cell–Cell Adhesion In Vitro and In Vivo [SUBSECTION] To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\\nTo study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\\n[SUBTITLE] DSG2-W2A Mutant Mice Resemble the Phenotype of ACM [SUBSECTION] Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\\nGiven the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\\n[SUBTITLE] Integrin-β6 Expression Is Altered in ACM Patient and DSG2-W2A Mouse Samples [SUBSECTION] Next, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\\nNext, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\\n[SUBTITLE] DSG2-W2A Mutant Hearts Present With Structurally Impaired ICDs [SUBSECTION] This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\\nThis result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\\n[SUBTITLE] Integrin-αV/β6 Is Activated in DSG2-W2A Mutant Hearts [SUBSECTION] In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\\nIn contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\\n[SUBTITLE] TGF-β Signaling Is Upregulated in DSG2-W2A Mutant Hearts [SUBSECTION] ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\\nITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\\n[SUBTITLE] Fibrosis in DSG2-W2A Mutant Animals Is Reduced by Inhibition of ITGAV/B6-Dependent Release of TGF-β [SUBSECTION] On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\\nOn the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\",\n \"To study the functional consequences of impaired DSG2 binding, we generated the W2A point mutation (DSG2-W2A) to abrogate the proposed interaction mechanism of the only desmoglein expressed in cardiomyocytes (Figure 1A). First, we investigated in a cell-free setup whether DSG2-W2A is indeed affecting the binding properties of DSG2 interaction by applying single molecule force spectroscopy. Here, molecules are coupled to the tip of an atomic force microscopy probe and the surface of a mica sheet by a polyethylene glycol linker. By measuring the deflection of the probe during repeated approach to and retraction from the surface, binding events can be assessed quantitatively. From these data, properties such as binding frequency, binding force, and bond half-life time can be determined. Constructs were generated for expression of wild-type (WT) and mutant (W2A) DSG2 extracellular domains fused to a human Fc fragment and tested for homotypic (WT:WT, W2A:W2A) and heterotypic (WT:W2A) interaction properties (Figure 1B). These experiments showed a significant reduction of the frequency of W2A:W2A as well as WT:W2A interactions compared with WT:WT (Figure 1C). The frequency of the remaining W2A bindings was comparable to probing Fc alone, which indicates that these are mostly nonspecific interactions. Homotypic WT interactions display a clear binding force peak, whereas the remaining W2A interaction forces were spread widely, again pointing to the nonspecificity of these interactions (Figure 1D). The lifetime of DSG2 interactions under zero force (τ0) was determined by fitting the binding forces detected at different loading rates,11 yielding τ0=1.088s with R=0.993. In contrast, no sufficient fit (R=0.509) was possible for W2A interactions (Figure 1E). Together, these data outline that the tryptophan swap is the major interaction mechanism of DSG2.\\nDSG2 (desmoglein-2) adhesion is mediated by tryptophan swap at position 2. A, Predicted interaction model of DSG2 extracellular domains by exchange of tryptophan residue at position 2 into a hydrophobic pocket of the opposing molecule. Cartoon 3D presentation of Protein Data Bank entry 5ERD9; tryptophan-2 is highlighted by ball and stick presentation. B, Schematic of single molecule force spectroscopy experiments. Recombinant extracellular domains (ECs) of wild-type DSG2 (DSG2-WT) or mutant DSG2 (DSG2-W2A) protein were coupled to a mica surface and atomic force microscopy cantilever by mean of a human Fc-tag (hFc) and a polyethylene glycol (PEG)20 linker and probed as indicated. C, Binding frequency of DSG2-W2A/DSG2-WT heterotypic and homotypic interactions at a pulling speed of 2 µm/s are shown. hFc served as control for nonspecific binding. *P<0.05, 1-way analysis of variance, Dunnett post hoc test. Each independent coating procedure with minimum 625 force curves is taken as biological replicate. D, Histogram of binding forces distribution with peak fit at a pulling speed of 2 µm/s corresponding to data in C. E, Determination of the bond half lifetime by the Bell equation11 of mean loading rates and binding forces analyzed from data of pulling speeds at 0.5, 1, 2, 5, and 7.5 µm/s. Average of values from 4 independent coating procedures with minimum 625 force curves each. R=R2, koff =off-rate constant, τ0=bond half lifetime under zero force. F, Dissociation assays to determine cell–cell adhesion were performed in CaCo2 cells (WT or DSG2 KO background) expressing DSG2-WT-mGFP or DSG2-W2A-mGFP constructs. mGFP empty vector served as control. *P<0.05, 1-way analysis of variance, Sidak post hoc test. Corresponding Western blot analysis confirmed effective expression of DSG2 constructs (*) versus the endogenous protein (arrow) in CaCo2 cells. α-tubulin (TUBA) served as loading control. G, Representative images of monolayer fragmentation from experiments in F. H, Dissociation assays in immortalized keratinocytes isolated from neonatal murine skin of the respective genotype. *P<0.05 or as indicated, Welch analysis of variance, Dunnett post hoc test. I, Macroscopic cardiac phenotype of DSG2-W2A mut/mut mice at age 15 weeks. J, Cardiac hypertrophy was analyzed as mean cross-sectional area of cardiomyocytes in hematoxylin & eosin–stained sections. Scale bar, 30 µm. *P<0.05, unpaired Student t test. K, Representative images of intercalated discs acquired by transmission electron microscopy, 3 mice per genotype. Orange asterisks mark intercellular widening, orange circle marks a ruptured junction. Scale bar, 1 µm.\\nTo determine the effect of the W2A mutation on a cellular level, DSG2-WT and DSG2-W2A constructs were stably expressed in the epithelial cell line CaCo2 either in a WT or DSG2-deficient background (CaCo2[WT] or CaCo2[DSG2 KO]).12 Cell–cell dissociation assays were performed, in which a confluent cell monolayer is detached and exposed to defined mechanical stress. Expression of DSG2-W2A in CaCo2(WT) cells significantly reduced intercellular adhesion as indicated by an increased number of fragments (Figure 1F and 1G). Whereas expressing DSG2-WT in the CaCo2(DSG2 KO) line significantly rescued the impaired intercellular adhesion in response to DSG2 loss, the expression of DSG2-W2A had no effect on fragment numbers.\\nBecause these data demonstrate the DSG2 tryptophan swap to be the central adhesive mechanism, we next generated a CRISPR/Cas9-based knock-in mouse model for DSG2-W2A to investigate the consequences of impaired DSG2 adhesion in vivo (Figure S1). In line with the studies in human cells, dissociation assays in keratinocyte cell lines generated from DSG2-W2A pups revealed reduced cell–cell adhesion in heterozygous (mut/wt) and homozygous (mut/mut) mutants (Figure 1H). Moreover, DSG2-W2A mice presented with a pronounced cardiac phenotype. Adult mut/mut mice demonstrated ventricular deformation, fibrotic and calcified areas, and hypertrophic cardiomyocytes (Figure 1I and 1J). In line with reduced intercellular adhesion of cardiomyocytes, transmission electron microscopy revealed disturbed ICDs with widened intercellular space and occasionally completely ruptured junctions in mut/mut hearts (Figure 1K).\\nA cardiac phenotype was also detectable during embryogenesis. Here, mating of heterozygous mice revealed a reduction of mut/mut offspring to 2.8% compared with the expected Mendelian ratio of 25%. Embryo dissections showed loss of the mut/mut animals between E12 and E14 (Figure S2A). The mut/mut embryos at day E12.5 appeared macroscopically pale with accumulation of blood in the cardiac area although the heart was still beating (Figure S2B). Cells suggestive for blood precursors were detectable in the pericardial space (Figure S2C). This finding suggests a rupture of the cardiac wall leading to pericardial bleeding and loss of mut/mut animals.\\nTogether, these in vitro and in vivo data outline an essential role of the DSG2 tryptophan swap for cell–cell adhesion and demonstrate severe pathologies associated with impaired DSG2 interaction.\",\n \"Given the cardiac affection in mutant mice, the notion that a majority of patients with ACM are reported with mutations in desmosomal genes,2 and that mutations specifically of W2 have already been associated with ACM,10 we examined homozygous and heterozygous mice including both sexes with regard to histologic and clinical ACM measures in an age range of 12 to 80 weeks.\\nHistologic analysis showed cardiac fibrosis, a major hallmark of ACM, in the myocardium of the right ventricle (RV) and left ventricle (LV) in mut/mut animals (Figure 2A through 2D), which did not profoundly increase over time. In contrast, heterozygous animals demonstrated a milder phenotype, with a fraction of animals (36%) starting to develop fibrosis in the RV around 6 months, whereas LV was unaffected. To better address differences over time, we pooled animals into groups with a mean age of 4 and 6 months for homozygous and 6 and 12 months for heterozygous mice and performed functional analyses by echocardiography and ECG parallel to histologic evaluation. Moreover, we separated 12-month-old heterozygous animals according to RV fibrosis into a group without and with fibrosis (>10% fibrotic area). Mut/mut animals showed impaired RV systolic function at both time points with reduced tricuspid annular plane systolic excursion as well as fractional shortening (Figure 2E through 2G and Table S1). LV systolic impairment with reduction of the ejection fraction and corresponding parameters worsened over time and became significant after 6 months, although no increase in fibrosis was detectable (Figure 2D and 2H and Table S1). In mut/wt mice, RV fractional shortening was significantly reduced in 12-month-old animals with fibrosis. In ECG recordings of mut/mut mice for 30 minutes under anaesthesia, we noted deformation of the QRS complex with elongated QRS interval and reduced amplitude of the S peak with effects more pronounced in the older group; heterozygous animals with RV fibrosis showed similar effects or trends (Figure 2I through 2K and Table S1). Because the T wave was not clearly detectable in mutants, the J peak amplitude as indicator for early repolarization was determined. This was reduced in the mut/mut groups with same trend for fibrotic mut/wt hearts (Figure 2L). Parallel to these depolarization and repolarization abnormalities, we also noted the occurrence of ventricular arrhythmia, ranging from single premature ventricular contractions (PVCs) to multiple, multifocal PVCs and nonsustained ventricular tachycardia in 1 mouse (Figure 2M through 2O). The fraction of homozygous and heterozygous animals with PVCs and the frequency of these events correlated with age and fibrosis. Moreover, a substantial fraction (16%) of mut/mut animals was found dead during the observation period, which strongly suggests the occurrence of malignant arrhythmia with sudden death (Figure 2P). Investigating sex-related differences, no effect was observed on the extent of fibrosis or echocardiographic parameters. However, in mut/wt mice, PVCs were found in males only and male mut/mut mice were more prone to premature sudden death compared with females (Figure S3).\\nDSG2 (desmoglein-2)–W2A mutant mice develop characteristics of arrhythmogenic cardiomyopathy. A, Cardiac fibrosis detected by picrosirius red collagen staining with representative images of sections from 6-month-old DSG2-W2A mut/mut and 12-month-old wt/wt and mut/wt animals. Scale bar = 1 mm. B through D, Corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Continuous lines indicate the simple linear regression between age and area of collagen. Hearts with >10% of collagen in the RV (gray dotted line) were defined as “with fibrosis.” Each dot represents 1 animal. *P<0.05, mixed effects analysis with LV and RV matched per animal, Sidak post hoc test. Lines indicate median and quartile values. E, Representative echocardiography images for measurements of the tricuspid annular plane systolic excursion (TAPSE) and LV function in 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Left side: 2D images from apical 4-chamber view for TAPSE and parasternal short axis view for LV. M-mode tracings on the right were performed along the line indicated on the left. Scale bars, white, 2 mm; black, 100 ms. Corresponding analysis of (F) TAPSE, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; (G) fractional shortening of the RV, *P<0.05, Kruskal-Wallis test with Dunn post hoc test; and (H) ejection fraction of the LV, *P<0.05, 1-way analysis of variance, Sidak post hoc test. I, ECG recoded in lead II with representative curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Definition of respective peaks is indicated in the wt/wt curve. Scale bars, vertical, 0.5 mV; horizontal, 50 ms. Corresponding analysis of (J) QRS interval, *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test; (K) amplitude of the S peak, *P<0.05 or as indicated, 1-way analysis of variance, Sidak post hoc test; and (L) amplitude of the J peak (early repolarization), *P<0.05 or as indicated, Kruskal-Wallis test with Dunn post hoc test. M, Ventricular arrythmia detected by ECG during 30 minutes of baseline measurements with example curves from 6-month-old mut/mut and 12-month-old wt/wt and mut/wt animals. Asterisks mark premature ventricular contractions (PVCs); *nsVT indicates a nonsustained ventricular tachycardia detected in 1 mut/mut animal. Scale bars, vertical, 0.5 mV; horizontal, 1 s. Corresponding analysis of (N) percentage of mice presenting with PVCs. Values in bars indicate corresponding absolute number of mice with PVC (colored bars) compared with total number of mice evaluated (empty bar) and (O) PVC burden depicted as number of PVCs per minute, *P<0.05, Kruskal-Wallis test with Dunn post hoc test. The black arrow indicates the animal presenting with nonsustained ventricular tachycardia. P, Kaplan-Meier survival plot of DSG2-W2A mice from an analysis period of 3 years. Vertical lines indicate dropouts because of unrelated elimination (end of experiment, breeding, injuries). Values indicate corresponding absolute number of mice with sudden death compared with total number of mice evaluated. *P<0.05, Gehan-Breslow-Wilcoxon test. Box with color indications of respective groups in the middle applies to the entire figure.\\nIn summary, mutant animals show histopathologic features with echocardiography and ECG abnormalities similar to patients with ACM. Within the limits of a murine model, both genotypes fulfil the Padua criteria13 to establish the diagnosis of ACM (Table S2). Here, DSG2-W2A mut/mut animals resemble the phenotype of biventricular ACM, whereas in mut/wt mice, the phenotype is milder with variable penetrance and age-dependent occurrence only in the RV. According to the Padua criteria,13 the heterozygous genotype recalls the characteristics of a right-dominant ventricular ACM, as long as fibrosis was present. Together, these data demonstrate that loss of desmosomal adhesion is sufficient to induce an ACM phenotype and that the DSG2-W2A model may resemble 2 different variants of the disease.\",\n \"Next, we applied this mouse model to investigate mechanisms leading to ACM. RNA sequencing was performed from ventricles of mut/mut and wt/wt hearts both before onset of fibrosis (age 5 days) and when the biventricular fibrosis and ACM phenotype was established (after 9 weeks). First, we compared sequencing data from 9-week-old mice with the top differentially expressed genes derived from published transcriptomic data sets of ACM patient hearts (GEO database: GSE107157/GSE10748014 and GSE2981915). Murine samples clustered according to their genotype and mut/mut hearts resembled the gene expression pattern of patients (Figure S4), further supporting the observation that mutant mice faithfully reproduce major aspects of the disease.\\nTo identify common genes deregulated during ACM pathogenesis, we compared the differentially expressed genes from patient data with the results from mutant mice. Here, we included data from 5-day-old mice to identify changes already present before onset of secondary effects attributable to fibrosis. Integrin-β6 (Itgb6) was the only transcript consistently deregulated in all data sets (Figure 3A). Its common downregulation in patients with ACM and mutant mice on RNA levels was further confirmed for heterozygous DSG2-W2A mutants compared with WT (Figure 3B). However, protein levels were unaltered in mutant heart samples (Figure 3C). Immunostaining revealed that, in mutant hearts, ITGB6, which mainly localized to a compartment suggestive for the transverse tubules and costameres in wt/wt animals, was enriched at the ICD (Figure 3D and E). Together, these data demonstrate altered ITGB6 mRNA expression as a common feature in ACM patient and DSG2-W2A hearts and an enrichment of the protein at the ICD in mutant murine samples.\\nIntegrin-β6 is deregulated in DSG2 (desmoglein-2)–W2A mutants. A, Venn diagram of significantly altered genes from indicated arrhythmogenic cardiomyopathy (ACM) patient data sets (ACM versus healthy control) and DSG2-W2A mice at the age of 5 days and 9 weeks (mut/mut versus wt/wt) highlighting integrin-β6 (Itgb6) as only overlapping gene with same direction of expression in all data sets. Numbers indicate the amount of overlapping genes for the respective overlays. B, RNA expression of Itgb6 analyzed by means of reverse transcription quantitative polymerase chain reaction in adult DSG2-W2A mouse hearts. *P<0.05, unpaired Student t test versus wt/wt. Gapdh served as reference gene. C, Representative Western blot and respective analysis of band intensity of integrin-β6 (ITGB6) in DSG2-W2A hearts. GAPDH served as loading control. 1-way analysis of variance, Dunnett post hoc test. D and E, Immunostaining of ITGB6 (red in overlay) in DSG2-W2A hearts with corresponding analysis of staining intensity in cardiomyocytes in total and ratio of staining intensity at the intercalated disc (ICD) area (orange arrowheads) versus cardiomyocytes’ cytosolic area. DSP (desmoplakin; cyan) marks ICDs, DAPI (blue) nuclei and F-actin (green) the sarcomere system. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Scale bars, 25 µm. *P<0.05; left graph: Kruskal-Wallis test with Dunn post hoc test; right graph: 1-way analysis of variance, Dunnett post hoc test. Box with color indications of respective groups applies to the entire figure.\",\n \"This result together with the transmission electron microscopy data (Figure 1K) suggest structural changes of the ICD in mutant hearts, which we addressed further. Using DSP as marker, we reconstructed the outlines of ICDs in wt/wt and mut/mut hearts from 3D stacks captured with structured illumination microscopy. This analysis yielded an increase in ICD volume, in particular attributable to an enhanced width between 2 adjacent cells (Figure 4A). The majority of desmosomal molecules as well as N-cadherin were regularly distributed at mutant ICDs (Figure S5A and S5B). This was corroborated by RNA sequencing data from 5-day-old and 9-week-old animals, which showed no consistent changes of desmosomal, adherens, gap, or tight junction molecules (Figure S5C through S5E). The only adhesion molecule reduced globally was DSG2 (Figure S5A and S5B). Although DSG2 was still present at the ICD, detailed analysis by structured illumination microscopy revealed reduced number and volume of DSG2 signals (Figure 4B). Moreover, fluorescence recovery after photobleaching experiments in neonatal murine cardiomyocytes transduced with DSG2-WT-mGFP or DSG2-W2A-mGFP showed a higher mobility of the mutant protein at cardiomyocyte junctions (Figure 4C), indicating reduced membrane stability of DSG2-W2A. To further investigate the possibility that the stability and integrity of ICD molecules is compromised, we performed Triton-X-100 separation assays for proteins with unaltered ICD localization in mutant hearts. Cytosolic and unanchored membrane proteins are found in the Triton-X-100 soluble fraction, whereas cytoskeleton-bound molecules are mainly detectable in the nonsoluble fraction. Both desmosomal molecules (PG, PKP2) as well as N-cadherin showed increased levels in the soluble pool, indicating reduced cytoskeletal anchorage and impaired ICD integrity in mutants (Figure 4D). Moreover, the gap junction molecule connexin-43, which is required for electrical coupling of cardiomyocytes, was delocalized to the lateral membrane (Figure 4E), which was described as a feature of disrupted ICDs.8\\nDisrupted ICD structure in DSG2 (desmoglein-2)–W2A mutant mice. A, Representative z-stack reconstruction of segmented intercalated discs (ICDs) in top and side view acquired with structured illumination microscopy. Overlay of analyzed ICD volume is shown in gray. ICDs are marked by DSP (desmoplakin; magenta). Scheme on top presents segmented area and pictograms on the right display the respective angle of view. Corresponding analysis of ICD volume and width of ICD between adjacent cardiomyocytes below. Scale bar, 5 µm. *P<0.05, Mann-Whitney test. Each dot represents the value of 1 ICD from 4 mice per genotype. B, Representative images of DSG2 (magenta) and filamentary actin (f-actin; white) z-stacks acquired by structured illumination microscopy and presented as maximum intensity projection. Lower row shows color-coded height projection of DSG2 signals in z-stack after signal thresholding as performed for analysis. Related analysis of DSG2 signal volume and number per ICD length is shown below. Pictograms on the right display the respective angle of view. Scale bar, 5 µm. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the value of 1 ICD from in total 3 mice per genotype. C, Fluorescence recovery after photobleaching analysis of DSG2–wild type (WT) and DSG2-W2A-mGFP fusion proteins at the cell–cell junction of neonatal cardiomyocytes with representative intensity kymographs of bleached areas on the left. Time point 0 = bleach as indicated by the black arrow. Analysis of the mobile fraction of the indicated mGFP-fusion proteins is shown on the right. *P<0.05, unpaired Student t test, with Welch correction. Each dot represents the mean value of 1 heart from in total 3 isolations. D, Representative triton-X-100 assay immunoblot with separation of a soluble (sol), noncytoskeletal bound protein fraction from a nonsoluble (non-sol), cytoskeletal-anchored fraction and corresponding analysis shown below. PG (plakoglobin), PKP2 (plakophilin-2), and N-cadherin (NCAD) were analyzed. Intensity of proteins was normalized to the total amount of protein detected by ponceau staining. GAPDH and DSP served as separation control. *P<0.05 or as indicated, unpaired Student t test (PG, NCAD, PKP2) or Mann-Whitney test (DSP). Each dot represents the result from 1 mouse. E, Immunostaining of connexin-43 (CX43; red in overlay) in DSG2-W2A hearts. NCAD (yellow) marks ICDs, DAPI (blue) nuclei, and F-actin (green) the sarcomere system. Orange arrowheads mark ICD, pink arrowheads highlight lateralization of CX43 staining. Scale bars, 25 µm. Images representative for 5 mice per genotype. Box with color indications of respective groups applies to the entire figure.\\nThese data together with the transmission electron microscopy results demonstrate structurally severely altered ICDs in response to the DSG2-W2A mutation and suggest a molecular basis leading to the functional alterations detectable by echocardiography and ECG.\",\n \"In contrast to DSG2, we detected ITGB6 to be increased at the ICD (Figure 3D and 3E), which was confirmed by analysis of structured illumination microscopy images (Figure 5A). ITGB6 needs to heterodimerize with integrin-αV (ITGAV) in order to be activated and bind to the extracellular matrix.16–18 Thus, we analyzed the expression of the counterpart ITGAV, which was significantly upregulated on protein level and increased at the ICD and costameres of mutant hearts (Figure 5B and 5C). In contrast, the localization and intensity of integrin-β1, as classical representative of the integrin group, was not altered (Figure S6). Staining with an antibody specifically recognizing the heterodimer of ITGAV/B6 revealed increased amount of dimer formation at the ICD and costameres in mutant mice (Figure 5D). Moreover, increased ITGAV/B6 staining intensity was also detectable in an ACM patient sample with DSP mutation (Figure 5E). To stabilize active integrins, recruitment of the cytoskeletal adapters talin and vinculin is required.19 Accordingly, the expression of talin-2 was increased in DSG2-W2A mut/mut myocardium, with vinculin being enriched at the ICD (Figure 5F and 5G). This effect is similar to what was shown before in response to increased intracellular traction forces on junctions.20 Thus, these data suggest that in response to a dysfunctional, nonadhesive ICD with impaired cytoskeletal anchorage, ITGAV/B6 dimers are recruited and activated at the compromised ICD as well as at the costamere region.\\nIncreased activation of ITGB6/AV at ICDs of DSG2 (desmoglein-2)–W2A mutant mice. A, Representative intercalated disc (ICD) reconstruction from z-stacks of integrin-β6 (ITGB6; green) and DSP (magenta) immunostaining acquired by structured illumination microscopy and presented as maximum intensity projection. Related analysis of ITGB6 signal volume and number per ICD length is shown below. Scale bar, 5 µm. *P<0.05 or as indicated, unpaired Student t test with Welch correction (left graph) and Mann-Whitney test (right graph). Each dot represents the value of 1 ICD from in total 4 mice per genotype. B, Representative Western blot and analysis of band intensity of integrin-αV (ITGAV) in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test with Welch correction. C, Immunostaining of ITGAV in DSG2-W2A hearts on the right with respective analysis on the left. DAPI (blue) marks nuclei. Lower row shows an overview image of a fibrotic area in mut/mut hearts. Dotted orange line marks the edge of fibrotic area. Scale bars: upper rows, 20 µm; lower row, 50 µm. *P<0.05, unpaired Student t test. D, Immunostaining of ITGAV/B6 heterodimer in DSG2-W2A mutant hearts with respective analysis of staining intensity on the right. Cyan rectangle marks zoomed area on the right. Scale bars: overview, 50 µm; insert, 10 µm. *P<0.05, unpaired Student t test. E, Representative immunostaining images of ITGAV/B6 heterodimer staining (red) in a patient with arrhythmogenic cardiomyopathy (ACM; DSP-E952X, heterozygous) and healthy control sample. F-actin (green) stains the sarcomere system. For the patient with ACM, 4 different tissue samples were analyzed and compared with 2 tissue samples from 2 healthy controls. Scale bar, 20 µm. F, Immunostaining of VCL (vinculin; magenta) and TLN2 (talin-2; red) in DSG2-W2A hearts on the right with respective analysis of the mean staining intensity in cardiomyocytes and ratio of the staining intensity at the ICD versus cytosolic area on the left. DAPI (blue) marks nuclei. F-actin (green) stains the sarcomere system. Scale bar, 25 µm; *P<0.05, unpaired Student t test. G, Representative Western blot and analysis of band intensity of VCL and TLN2 in DSG2-W2A hearts. GAPDH served as loading control. *P<0.05, unpaired Student t test for TLN2, Mann-Whitney test for VCL. Box with color indications of respective groups applies to the entire figure.\",\n \"ITGAV/B6 dimers have the ability to activate the profibrotic cytokine transforming growth factor–β (TGF-β) by binding and removal of the latency-associated peptide.17 Gene set enrichment analyses revealed an upregulation of genes associated with TGF-β signaling in both 9-week-old mutants as well as patients with ACM (Figure 6A). Moreover, direct targets of receptor-regulated SMADs involved in TGF-β signaling21 were upregulated in DSG2-W2A mutant hearts (Figure 6B). Accordingly, increased amounts of nuclear SMAD2/3 phosphorylated at the activation sites S465/S467 or S423/S425, respectively, were detected in these hearts (Figure 6C). Increased levels of pSMAD2/3 were found not only in fibroblasts but also in cardiomyocytes, mainly adjacent to fibrotic areas. These data indicate upregulation of the profibrotic TGF-β pathway and its downstream targets, which include extracellular matrix proteins such as collagens, laminin, or fibronectin.\\nElevated transforming growth factor–β signaling in DSG2 (desmoglein-2)–W2A hearts as result of ITGAV/B6 activity. Barcode plots of gene set enrichment analysis of (A) the KEGG_TGF_BETA_SIGNALING_PATHWAY data set (systematic name: M2642)41 and (B) genes directly regulated by receptor-regulated SMADs (R-SMADs; includes SMAD1/2/3/5/9) as published in the TRRUST database21 in 9-week-old DSG2-W2A (mut/mut vs wt/wt) or arrhythmogenic cardiomyopathy (ACM) patient data set 1 (ACM versus healthy control; GEO: GSE107157/GSE107480). Indicated P values are calculated by function cameraPR of the R package limma. C, Immunostaining of phosphorylated SMAD2/3 (magenta, S465/S467 or S423/S425, respectively) in sections of DSG2-W2A hearts and related analysis of nuclear staining intensity. Nuclei are stained with DAPI (blue), cardiomyocytes are marked with f-actin (green). The dotted orange line marks the edge of the fibrotic area. Scale bar, 50 µm. *P<0.05, unpaired Student t test. D, Schematic of experimental setup for integrin-αVβ6 (ITGAV/B6) blocking experiments in cardiac slice culture with related results in E. Icons are derived from BioRender. E, Reverse transcription quantitative polymerase chain reaction analysis of expression of genes downstream of transforming growth factor–β (TGF-β) signaling in cardiac slice cultures treated with inhibiting anti-ITGAV/B6 (1:15) or 10 µmol/L GW788388, an inhibitor of TGF-β receptor I, for 24 hours. *P< 0.05, paired Student t test versus indicated control condition. F, Schematic of experimental setup for in vivo ITGAV/B6 blocking experiments by injection of 40 mg/kg EMD527040 (EMD) intraperitoneally daily. DMSO was applied as vehicle control. Icons are derived from BioRender. G, Reverse transcription quantitative polymerase chain reaction expression analysis of genes downstream of TGF-β signaling in hearts of mice treated with EMD or respective amount of DMSO for 10 days. *P<0.05, unpaired Student t test versus DMSO. H and I, Cardiac fibrosis detected by picrosirius red collagen staining with representative images and corresponding analysis of the area of collagen in the right ventricle (RV) and left ventricle (LV). Lines indicate littermates. Each dot represents 1 animal. *P<0.05 or as indicated, grouped 2-way repeated-measures analysis of variance with LV/RV and experimental pairs matched, Sidak post hoc test. Scale bar, 1 mm. J through L, ECG recoded in lead II with representative curves shown in J. Corresponding analysis of R, S, and J peak amplitude and QRS interval, *P<0.05 or as indicated. Paired Student t test. Lines indicate littermates. Each dot represents 1 animal.\",\n \"On the basis of the known profibrotic role of TGF-β in cardiomyopathies2 and the established function of ITGAV/B6 dimers to activate TGF-β,17 we investigated whether inhibition of ITGAV/B6 was efficient to reduce fibrosis in DSG2-W2A mice. Cardiac slice cultures were generated from the ventricles of heterozygous mice at the age of 40 to 50 weeks and treated for 24 hours with anti-ITGAV/B6, which was shown to neutralize the function of the dimer.22 In comparison, slices were treated with the TGF-β receptor I inhibitor GW788388 to directly block TGF-β signaling (Figure 6D). Significant downregulation in the expression of the profibrotic molecules collagen-I type-α1 (Col1a1), laminin subunit-γ2 (Lamc2), metalloproteinase inhibitor-1 (Timp1), and ID-2 (Id2) was induced in the mutant mice in response to inhibition of ITGAV/B6 (Figure 6E). Other extracellular matrix proteins such as fibronectin (Fn1) or collagens (Col3a1, Col1a2) showed the same tendency. All of these fibrotic markers were significantly upregulated in adult mutant hearts as detected by RNA sequencing (Figure 6A and 6B). Direct inhibition of TGF-β led to a similar downregulation of these markers as the blocking antibody.\\nTo investigate the in vivo relevance of this pathway, we applied the small molecule EMD527040 (EMD), which is an established inhibitor of ITGAV/B6-dependent TGF-β release, to the DSG2-W2A mouse model.23 For these experiments, we chose wt/mut mice because they reflect the heterozygosity most commonly found in patients with ACM and develop the phenotype during adulthood, when compound application is feasible. Littermates between the ages of 28 and 38 weeks were sex-matched and injected intraperitoneally daily for 10 days with EMD or DMSO control (Figure 6F). In this short-time approach, myocardial samples of EMD-treated animals showed a similar reduction of profibrotic markers as reported for slice cultures (Figure 6G). Having established an effective dose, we performed in vivo studies for a duration of 60 days during the time period when RV fibrosis establishes. At the end of the treatment period, ECG and histologic analysis were performed. EMD-treated mice demonstrated reduced RV fibrosis (Figure 6H and 6I). Moreover, EMD animals showed a mitigation of ECG abnormalities, with significantly higher S amplitude and trends towards an increased R amplitude and shorter QRS interval compared with the vehicle-treated control.\\nThese experiments link ITGAV/B6 dimerization in response to abrogated DSG2 binding and ICD dysfunction to TGF-β–dependent fibrosis generation. The pilot inhibitor studies suggest ITGAV/B6 as a promising treatment target to ameliorate the ACM phenotype.\",\n \"In this study, we generated a knock-in mouse model with defective binding function of the adhesion molecule DSG2, which demonstrated that impaired desmosomal adhesion is sufficient to induce a phenotype mimicking the characteristics of ACM. Our data suggest a cascade of defective desmosomal adhesion, disrupted ICD structure, and subsequent activation of ITGAV/B6 with profibrotic TGF-β signaling as important underlying mechanisms leading to this phenotype (Figure 7). Moreover, our pilot study indicates a beneficial effect of ITGAV/B6 inhibition by EMD treatment with regard to fibrosis and several ECG parameters, suggesting that this pathway can be targeted successfully by drug treatment.\\nSchematic conclusion of data. DSG2 (desmoglein-2) –W2A mutation with loss of desmosomal adhesion leads to impaired intercalated disc (ICD) structure with deregulation of integrin-β6 (ITGB6) and enhanced heterodimerization with integrin-αV (ITGAV). The dimer efficiently binds to the extracellular matrix and activates transforming growth factor–β (TGF-β) by removal of the latency-associated peptide (LAP). Active TGF-β can then induce profibrotic downstream signaling by means of SMAD molecules. In our experiments, this cascade was blocked by different approaches to inhibit ITGAV/B6.\\n[SUBTITLE] Defective Desmosomal Adhesion by DSG2-W2A Mutation Is Inducing an ACM Phenotype [SUBSECTION] Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\\nBecause of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\\n[SUBTITLE] DSG2-W2A Mice as Model to Study ACM [SUBSECTION] Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\\nOur results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\\n[SUBTITLE] DSG2-W2A Leads to Dysfunctional ICDs and ITGAV/B6 Rearrangement [SUBSECTION] DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\\nDSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\\n[SUBTITLE] Activation of TGF-β Signaling by ITGAV/B6 as Potential Therapeutic Target in ACM [SUBSECTION] The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.\\nThe ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.\",\n \"Because of mutations mainly affecting desmosomal genes and evidence of disrupted ICDs, the hypothesis of dysfunctional desmosomes with loss of cell–cell adhesion as a central pathologic step was adopted in the field.2 However, experimental data on this topic are contradictory.6–8 We show that W2 of DSG2 is central for binding and intercellular adhesion in vitro and in vivo. Disrupted desmosomal adhesion is sufficient to induce an ACM phenotype fulfilling the Padua criteria,13 including ventricular fibrosis, depolarization and repolarization abnormalities, arrhythmia, and impaired ventricular function, which are used to establish the diagnosis. In line with this, 3 mutations were described in patients with ACM directly affecting the DSG2 binding mechanism by exchange of the tryptophan with a serine, leucine, or arginine (ClinVar data bank: VCV000199796, VCV000044282, VCV000420241).10 This indicates a disruption of the tryptophan swap independent from the substituting amino acid and supports defective desmosomal adhesion as important factor in ACM.2,24 Moreover, different published DSG2 mutant mouse models show a phenotype resembling the features of DSG2-W2A mutants. These include a DSG2 mutant in which parts of the 2 outermost DSG2 extracellular domains were deleted (but leaving W2A intact),25 2 mouse lines with loss of DSG2,25,26 and mice overexpressing a patient mutation (DSG2-N271S),27 which also showed that ICD structural aberrations precede functional abnormalities and fibrosis generation. These data fuel the hypothesis that disruption and rearrangement of the ICD in response to impaired adhesion between cardiomyocytes is a crucial initial step, at least in case of mutations in desmosomal molecules. Other mechanisms described in patients or translational models, such as aberrant WNT or Hippo/YAP signaling, or immune cell infiltrations,2,28 may be secondary responses and in part even represent adaptive attempts to rescue the functional consequences of such severe structural aberrations.\",\n \"Our results suggest the DSG2-W2A model as a valuable tool to study ACM mechanisms. Within the limitations of a murine model, it reproduces a majority of features found in patients with ACM (Table S2). In homozygous animals, both ventricles are affected from early on in life; heterozygous animals develop a milder phenotype, with fibrosis occurring only in the RV. Whether this milder phenotype would extend to the LV over time or whether a gene dose effect underlies the structural differences is unclear. LV and RV remodel differently in response to loading and injury29 and it is possible that the RV is less able to compensate a partial reduction of cardiomyocyte cohesion, in contrast to the LV. A fraction of mut/wt animals did not develop the phenotype at all within the observation period of up to 80 weeks. Thus, the model might help clarify why patients with the same mutation (ie, in the same family) develop the disease with variable penetrance.1 Similar to patients,1 male mutant mice appear to experience more arrhythmia and sudden death than do female mice.\\nWe can interpret from the combined histology, echocardiography, and ECG data that the changes in heart function are secondary to fibrosis generation. This is indicated by the notion that functional measures (eg, ejection fraction, QRS interval) worsened over time in homozygous mutants, whereas no increase was detectable for fibrosis. Moreover, functional changes in mut/wt animals were only observed when fibrosis was present. This is in line with a patient cohort with DSP mutations, in which LV fibrosis preceded systolic dysfunction.30 A different study in DSG2-N271S-expressing ex vivo paced hearts demonstrated conduction velocity impairments before onset of fibrosis.27 More longitudinal studies using different mutational backgrounds and sensitive detection methods (eg, cardiac magnetic resonance imaging and ECG under exercise) are necessary to define the role of fibrosis as cause or consequence of functional changes.\\nAs with all murine models, the DSG2-W2A line has limitations in its ability to fully recapitulate the human disease phenotype. Mutant mice do not show pronounced replacement with adipose tissue, which is a common characteristic in patient heart samples, but in general is not fully modeled in mice.24 Moreover, at least in ECGs under anesthesia, not all mutant animals develop arrhythmia, even if fibrosis is present. However, in contrast to most other ACM models, this knock-in model has the advantage of a single amino acid exchange under the endogenous promotor, thus reducing the possibility of secondary unwanted effects attributable to complete protein absence or overexpression. As in patients, the mutation is present in all cell types and not limited to cardiomyocytes.2,24\",\n \"DSG2-W2A led to a severely altered ICD structure consistent with impaired cytoskeletal attachment and ruptured junctions. Whereas these changes provide explanations for compromised intercellular adhesion, we also noted alterations in cell–matrix protein distribution. RNA sequencing before and after onset of the disease phenotype and a comparison with ACM patient datasets identified ITGB6 as commonly deregulated in patients with ACM and DSG2-W2A mice. Although we noted reduced mRNA levels in mutant hearts, the overall protein content was unaltered. This suggests pronounced posttranslational regulation of ITGB6, which was already demonstrated in skeletal muscle.31 In mutant hearts, ITGB6 was increased at the ICD together with elevated levels of ITGAV and increased heterodimerization of both molecules. By mechanical force exerted through ITGAV/B6, TGF-β is detached from the latency-associated peptide and can induce signaling by means of TGF-β receptors.17,32 In line with higher intracellular forces, talin-2 and vinculin were increased at ICDs, as binding of these molecules activates and stabilizes integrins. A mutual regulation of both adhesive compartments is well-established. As an example, upon loss of N-cadherin, integrins are activated and induce fibronectin deposition.33 A similar mechanism is conceivable in DSG2-W2A hearts, either by loss of DSG2 or because of reduced N-cadherin anchorage. So far, only limited data are available on the role of integrins in ACM. A recent study showed downregulation of integrin-β1D leading to ventricular arrhythmia.34 Furthermore, knockdown of PKP2 in HL-1 cardiomyocytes was described to deregulate focal adhesions, including integrin-α1.35\",\n \"The ITGAV/B6 heterodimer is described as one of the major activators of latent TGF-β1 and TGF-β3.17 Although TGF-β signaling is known as general driver of cardiac fibrosis36 and more specifically was implicated in ACM,2 to our knowledge no data are available on the role of ITGAV/B6 and their regulation of TGF-β signaling in cardiac fibrosis. Uncovering this pathway is of high interest as it offers the possibility to target TGF-β with reduced risk of severe side effects occurring in response to direct inhibition.37 We demonstrate a reduction of profibrotic gene expression under ACM conditions in response to ITGAV/B6 inhibition and our pilot study data suggest that a small molecule blocking ITGAV/B6-dependent TGF-β release diminished the generation of fibrosis. Similar approaches using this small molecule or neutralizing antibodies were shown to be protective in murine models of lung, liver, and biliary fibrosis.23,38,39 However, because of the heterogeneity of the phenotype in heterozygous animals, more detailed studies with larger sample sizes are required to further substantiate this finding.\\nWhether fibrosis generation is a contributor to the disease or a protective mechanism to ensure integrity of the heart under conditions of compromised adhesion is unclear. In case of the latter, pharmacologic inhibition of fibrosis generation might be detrimental. Nevertheless, fibrosis is a driver of arrhythmia generation,40 which is also supported by our results that ECG abnormalities were reduced by inhibition of fibrosis. Careful studies using different therapeutic approaches to reduce fibrosis in appropriate model organisms need to address this aspect in detail.\\nIn conclusion, we established a new mouse model phenocopying many aspects of ACM, uncovered a novel pathway of fibrosis induction, and identified an approach to target this mechanism with future implication as a potential therapeutic option.\",\n \"[SUBTITLE] Acknowledgments [SUBSECTION] The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\\nThe authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\\n[SUBTITLE] Sources of Funding [SUBSECTION] This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\\nThis project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\\n[SUBTITLE] Disclosures [SUBSECTION] Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\\nDr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\\n[SUBTITLE] Supplemental Material [SUBSECTION] Expanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\\nExpanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\",\n \"The authors thank Alain Brühlhart and the team from the Animal Facility; Dr Cinzia Tiberi (Center for Cellular Imaging and NanoAnalytics); Dr Alexia Loynton-Ferrand (Imaging Core Facility); Dr Diego Calabrese (Histology Core Facility); and Drs Mike Abanto and Beat Erne, P. Lorentz, and E. Bartoszek (Microscopy Core Facility), University of Basel, Switzerland; and Dr Christian Beisel and Philippe Demougin, Genomics Facility Basel (D-BSSE, ETH Zürich, and University of Basel), for support; Nicolas Schlegel (Department of Surgery, University Hospital Würzburg, Germany) for providing CaCo2 cells; Nikola Golenhofen (Institute of Anatomy, University of Ulm, Germany) for providing the Fc-His-pEGFP-N3 plasmid; and Anja Fuchs for technical assistance. Calculations were performed at the sciCORE scientific computing center at University of Basel.\",\n \"This project was supported by the German Research Council (SP1300-3/1 to Dr Spindler), the Swiss National Science Foundation (197764 to Dr Spindler), the Swiss Heart Foundation (FF21098 to Drs Schinner and Spindler), the Research Fund Junior Researchers, University of Basel (3BM1079 to Dr Schinner), and the Theiler-Haag Foundation (Dr Spindler). Dr Sheikh is supported by grants from the National Institutes of Health (grant HL142251) and Department of Defense (grant W81XWH1810380).\",\n \"Dr Sheikh is a cofounder and equity shareholder of Papillon Therapeutics Inc and is a consultant and equity shareholder of LEXEO Therapeutics Inc. The other authors have nothing to disclose.\",\n \"Expanded Methods\\nTables S1–S3\\nFigures S1–S6\\nReferences 42–48\",\n \"\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,"methods",null,null,null,"data-availability","results",null,null,"subjects",null,null,null,null,"discussion",null,null,null,null,null,null,null,"COI-Statement",null,"supplementary-material"],"string":"[\n null,\n null,\n \"methods\",\n null,\n null,\n null,\n \"data-availability\",\n \"results\",\n null,\n null,\n \"subjects\",\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"COI-Statement\",\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["cardiomyopathies","desmosomes","fibrosis","integrins","intercalated disc","transforming growth factors"],"string":"[\n \"cardiomyopathies\",\n \"desmosomes\",\n \"fibrosis\",\n \"integrins\",\n \"intercalated disc\",\n \"transforming growth factors\"\n]"}}},{"rowIdx":381,"cells":{"title":{"kind":"string","value":"End-stage renal disease incidence in a cohort of US firefighters from San Francisco, Chicago, and Philadelphia."},"pmid":{"kind":"string","value":"36268894"},"background_abstract":{"kind":"string","value":"Firefighters perform strenuous work in hot environments, which may increase their risk of chronic kidney disease. The purpose of this study was to evaluate the risk of end-stage renal disease (ESRD) and types of ESRD among a cohort of US firefighters compared to the US general population, and to examine exposure-response relationships."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"ESRD from 1977 through 2014 was identified through linkage with Medicare data. ESRD incidence in the cohort compared to the US population was evaluated using life table analyses. Associations of all ESRD, systemic ESRD, hypertensive ESRD, and diabetic ESRD with exposure surrogates (exposed days, fire runs, and fire hours) were examined in Cox proportional hazards models adjusted for attained age (the time scale), race, birth date, fire department, and employment duration."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The incidence of all ESRD was less than expected (standardized incidence ratio (SIR) = 0.79; 95% confidence interval = 0.69-0.89, observed = 247). SIRs for ESRD types were not significantly increased. Positive associations of all ESRD, systemic ESRD, and hypertensive ESRD with exposed days were observed: however, 95% confidence intervals included one."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"We found little evidence of increased risk of ESRD among this cohort of firefighters. Limitations included the inability to evaluate exposure-response relationships for some ESRD types due to small observed numbers, the limitations of the surrogates of exposure, and the lack of information on more sensitive outcome measures for potential kidney effects."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Aged","United States","Firefighters","Incidence","Chicago","Philadelphia","San Francisco","Medicare","Kidney Failure, Chronic"],"string":"[\n \"Humans\",\n \"Aged\",\n \"United States\",\n \"Firefighters\",\n \"Incidence\",\n \"Chicago\",\n \"Philadelphia\",\n \"San Francisco\",\n \"Medicare\",\n \"Kidney Failure, Chronic\"\n]"},"pmcid":{"kind":"string","value":"9828160"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Cohort description [SUBSECTION] The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\nThe previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\n[SUBTITLE] Exposure assessment [SUBSECTION] Exposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\nExposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\n[SUBTITLE] Ascertainment of vital status and ESRD [SUBSECTION] Vital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n\nVital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n\n[SUBTITLE] Analysis [SUBSECTION] The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\nThe incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"Table 1 shows the demographic characteristics of the full and restricted cohorts. Most members of the full cohort were male (96.3%) and White (78.6%). Over three‐fourths (78.1%) of the restricted cohort of male firefighters were also White. The mean employment duration for both cohorts was 21 years. At the end of the follow‐up, 38.2% of the full cohort and 25.3% of the restricted cohort were deceased.\nCharacteristics of the full and restricted cohorts\nAbbreviation: IQR, interquartile range.\nMale firefighters of a known race who were hired in 1950 or later and employed for at least 1 year.\nVital status at the end of follow‐up (December 31, 2014) or the date of diagnosis, whichever was earlier.\nCell size <11 or reporting the exact number allows a cell size <11 to be derived.\nAs shown in Table 2, the incidence of all ESRD was less than expected among the full (SIR = 0.79; 95% CI = 0.69–0.89; observed = 247) and restricted (SIR: 0.73; 95% CI = 0.63–0.85; observed = 177) cohorts. The median age at the time of onset was 69 years (interquartile range = 61–77) and 67 years (interquartile range: 59–73) among the full and restricted cohorts, respectively. SIRs for ESRD due to metabolic disease (full cohort: 2.84; 95% CI = 0.76–7.26; restricted cohort: 1.92; 95% CI = 0.22–6.94) and ESRD due to collagen vascular disease (full cohort: 1.18; 95% CI = 0.32–3.01; restricted cohort: 1.17; 95% CI = 0.24–3.42) were elevated, although the CIs for these ESRD types were wide and included one. The incidence of systemic ESRD and diabetic ESRD was less than expected among the full (systemic ESRD: SIR = 0.76; 95% CI = 0.65–0.88; diabetic ESRD: SIR = 0.66; 95% CI = 0.52–0.82) and restricted (systemic ESRD: SIR = 0.71; 95% CI = 059–0.84; diabetic ESRD: SIR = 0.63; 95% CI = 0.49–0.81) cohorts. The incidence of ESRD caused by glomerulonephritis was also less than expected among the full cohort (SIR = 0.61; 95% CI = 0.35–0.97). The SIR for ESRD caused by glomerulonephritis among the restricted cohort was 0.66 (95% CI = 0.36–1.10).\nSIRs for ESRD among the full and restricted cohorts, 1977–2014\nAbbreviations: AIDS, acquired immunodeficiency syndrome; CI, confidence interval; ESRD, end‐stage renal disease; NR, not reported (cell size <11 or reporting the cell size allows numbers <11 to be derived); Obs, observed; SIR, standardized incidence ratio.\nESRD due to AIDS, sickle cell disease, and hereditary nephropathy were omitted because no cases were observed.\nThe results of internal analyses among the restricted cohort using categorical models are shown in Table 3 (20‐year lag) and Supporting Information: Table S2 (10‐year lag). Results reported herein were obtained from the analyses with matching on employment duration, unless stated otherwise. HRs for all ESRD obtained from categorical models increased in the second and third quartiles of exposed days, but then decreased in the top quartile. HRs for systemic ESRD increased with each exposed‐day quartile in 20‐ and 10‐year lagged analyses. HRs for hypertensive ESRD also increased with the exposed‐day quartile in 20‐ and 10‐year lagged analyses with an HR of 4.51 (95% CI = 0.97–20.9) in the top quartile in 20‐year lagged analyses. HRs for diabetic ESRD increased with exposed‐day quartile in 10‐year lagged analyses only.\nESRD HRs by 20‐year lagged exposure surrogates obtained from categorical models\na\n\n\nAbbreviations: CFD, Chicago fire department; CI, confidence interval; dept, fire departments included in the analysis; ESRD, end‐stage renal disease; HR, hazard ratio; PFD, Philadelphia fire department; SFFD, San Francisco fire department.\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases with an approximately equal number of cases in each exposure quartile. The overall numbers of ESRD cases in the analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\nThe results of internal analyses among the restricted cohort using log‐linear models are shown in Table 4 (20‐year lag) and Supporting Information: Table S3 (10‐year lag). The HR for all ESRD was 1.11 (95% CI = 0.99–1.25) and 1.08 (95% CI = 0.98–1.18) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top exposure decile was excluded. In analogous linear models including person‐time in the top decile of exposure, the HR for all ESRD was 1.05 (95% CI = 0.96–1.16) and 1.07 (0.99–1.16) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively.\nESRD HRs by 20‐year lagged exposure surrogates obtained from log‐linear models\na\n\n\nAbbreviations: CI, confidence interval; ESRD, end‐stage renal disease; HR, hazard ratio.\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases. The overall numbers of ESRD cases in analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\nExposed days were available for Chicago, Philadelphia, and San Francisco firefighters; fire runs for Chicago and Philadelphia firefighters; and fire hours for Chicago firefighters.\n1000 exposed days, 1000 fire runs, or 1000 fire hours.\nSystemic ESRD HRs per 1000 exposed days were 1.05 (95% CI = 0.95–1.18) and 1.07 (95% CI = 0.98–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was included and 1.07 (95% CI = 0.95–1.23) and 1.05 (95% CI = 0.95–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was excluded. Hypertensive ESRD HRs per 1000 exposed days were 1.09 (95% CI = 0.93–1.29) and 1.10 (95% CI = 0.96–1.27) in 20‐ and 10‐year lagged analyses, respectively. When person‐time in the top exposure decile was excluded, the hypertensive ESRD HRs per 1000 exposed days were 1.13 (95% CI = 0.93–1.40) and 1.09 (95% CI = 0.94–1.29) in 20‐ and 10‐year lagged analyses, respectively. In 10‐year lagged analyses, the diabetic ESRD HR per 1000 exposed days was 1.03 (95% CI = 0.92–1.17) and 1.03 (95% CI = 0.90–1.19), including and excluding person‐time in the top exposure decile, respectively."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Cohort description","Exposure assessment","Ascertainment of vital status and ESRD","Analysis","AUTHOR CONTRIBUTIONS","DISCLOSURE BY AJIM EDITOR OF RECORD","ETHICS APPROVAL AND INFORMED CONSENT","DISCLAIMER"],"string":"[\n \"INTRODUCTION\",\n \"Cohort description\",\n \"Exposure assessment\",\n \"Ascertainment of vital status and ESRD\",\n \"Analysis\",\n \"AUTHOR CONTRIBUTIONS\",\n \"DISCLOSURE BY AJIM EDITOR OF RECORD\",\n \"ETHICS APPROVAL AND INFORMED CONSENT\",\n \"DISCLAIMER\"\n]"},"other_sections_texts":{"kind":"list like","value":["Firefighters may be at increased risk of developing chronic kidney disease because they perform strenuous work in hot environments. Strenuous work with heat stress and dehydration may cause renal impairment from clinical or subclinical rhabdomyolysis, hyperuricemia and urate crystalluria, and hyperosmolality‐induced release of vasopressin and activation of the aldose reductase‐fructokinase pathway.\n1\n, \n2\n, \n3\n, \n4\n, \n5\n, \n6\n, \n7\n, \n8\n, \n9\n, \n10\n, \n11\n Increased concern about these potential heat‐related effects on kidney function has arisen due to an epidemic of chronic kidney disease among sugarcane workers in Central America.\n12\n Chronic kidney disease in these workers has often progressed to end‐stage renal disease (ESRD).\n3\n, \n13\n Chronic kidney disease has also been observed among other heat‐exposed workers and in other regions.\n4\n, \n12\n, \n14\n, \n15\n, \n16\n, \n17\n However, few data exist on the risk of kidney disease among firefighters.\nFirefighters are at risk of heat stress from exposure to heat from fires, ambient heat during hot weather, and metabolic heat generated by strenuous physical work.\n18\n, \n19\n The heavy, impermeable personal protective equipment structural firefighters wear adds to their metabolic workload and impairs evaporative cooling.\n19\n Increased core body temperature and heat‐related illnesses, including rhabdomyolysis and fatal heat stroke, have been documented among firefighters and trainees.\n20\n, \n21\n, \n22\n, \n23\n, \n24\n, \n25\n, \n26\n, \n27\n, \n28\n Firefighters are also at risk of dehydration,\n29\n which can contribute to kidney injury.\n30\n, \n31\n\n\nWe previously evaluated mortality and cancer incidence among a large National Institute for Occupational Safety and Health (NIOSH) cohort of US firefighters from the Chicago, Philadelphia, and San Francisco fire departments and found an increased risk of several cancers including kidney cancer compared to the US general population.\n32\n, \n33\n Mortality from chronic and unspecified nephritis and renal failure and other renal sclerosis, based on the underlying cause of death, was elevated among Philadelphia firefighters (standardized mortality ratio = 1.35; 95% confidence interval [CI] = 1.03–1.73) but not Chicago or San Francisco firefighters.\n33\n However, chronic kidney disease is usually not the underlying cause of death even when mentioned on the death certificate.\n34\n In this study, we evaluate ESRD incidence in the NIOSH firefighter cohort compared to the general population and relationships of ESRD with surrogates of exposure. Chicago and Philadelphia have humid continental climates, whereas San Francisco has a Mediterranean climate.\n35\n From 1981 to 2010, the average annual number of days the temperature reached 32.2°C was 17 in Chicago, 25 in Philadelphia, and 2 in San Francisco.\n36\n\n","The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.","Exposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.","Vital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n","The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.","Lynne E. Pinkerton conceived the study and wrote the first draft of the manuscript. Stephen Bertke conducted the analysis. Travis L. Kubale, James H. Yiin, and Robert D. Daniels acquired the data and assembled the cohort. Matthew M. Dahm led the assessment of the exposure surrogates. All authors participated in the interpretation and presentation of results, approved the final manuscript, and agreed to be accountable for all aspects of the work.","John Meyer declares that he has no conflict of interest in the review and publication decision regarding this article.","The study received approvals from the Institutional Review Boards of the National Institute for Occupational Safety and Health and the National Cancer Institute. Informed consent was waived for this records‐based study.","The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention."],"string":"[\n \"Firefighters may be at increased risk of developing chronic kidney disease because they perform strenuous work in hot environments. Strenuous work with heat stress and dehydration may cause renal impairment from clinical or subclinical rhabdomyolysis, hyperuricemia and urate crystalluria, and hyperosmolality‐induced release of vasopressin and activation of the aldose reductase‐fructokinase pathway.\\n1\\n, \\n2\\n, \\n3\\n, \\n4\\n, \\n5\\n, \\n6\\n, \\n7\\n, \\n8\\n, \\n9\\n, \\n10\\n, \\n11\\n Increased concern about these potential heat‐related effects on kidney function has arisen due to an epidemic of chronic kidney disease among sugarcane workers in Central America.\\n12\\n Chronic kidney disease in these workers has often progressed to end‐stage renal disease (ESRD).\\n3\\n, \\n13\\n Chronic kidney disease has also been observed among other heat‐exposed workers and in other regions.\\n4\\n, \\n12\\n, \\n14\\n, \\n15\\n, \\n16\\n, \\n17\\n However, few data exist on the risk of kidney disease among firefighters.\\nFirefighters are at risk of heat stress from exposure to heat from fires, ambient heat during hot weather, and metabolic heat generated by strenuous physical work.\\n18\\n, \\n19\\n The heavy, impermeable personal protective equipment structural firefighters wear adds to their metabolic workload and impairs evaporative cooling.\\n19\\n Increased core body temperature and heat‐related illnesses, including rhabdomyolysis and fatal heat stroke, have been documented among firefighters and trainees.\\n20\\n, \\n21\\n, \\n22\\n, \\n23\\n, \\n24\\n, \\n25\\n, \\n26\\n, \\n27\\n, \\n28\\n Firefighters are also at risk of dehydration,\\n29\\n which can contribute to kidney injury.\\n30\\n, \\n31\\n\\n\\nWe previously evaluated mortality and cancer incidence among a large National Institute for Occupational Safety and Health (NIOSH) cohort of US firefighters from the Chicago, Philadelphia, and San Francisco fire departments and found an increased risk of several cancers including kidney cancer compared to the US general population.\\n32\\n, \\n33\\n Mortality from chronic and unspecified nephritis and renal failure and other renal sclerosis, based on the underlying cause of death, was elevated among Philadelphia firefighters (standardized mortality ratio = 1.35; 95% confidence interval [CI] = 1.03–1.73) but not Chicago or San Francisco firefighters.\\n33\\n However, chronic kidney disease is usually not the underlying cause of death even when mentioned on the death certificate.\\n34\\n In this study, we evaluate ESRD incidence in the NIOSH firefighter cohort compared to the general population and relationships of ESRD with surrogates of exposure. Chicago and Philadelphia have humid continental climates, whereas San Francisco has a Mediterranean climate.\\n35\\n From 1981 to 2010, the average annual number of days the temperature reached 32.2°C was 17 in Chicago, 25 in Philadelphia, and 2 in San Francisco.\\n36\\n\\n\",\n \"The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\\n32\\n, \\n33\\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\",\n \"Exposure surrogates were calculated for the restricted cohort as described previously.\\n37\\n, \\n38\\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\\n38\\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\",\n \"Vital status was ascertained through 2016 as described previously.\\n32\\n, \\n33\\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\\n39\\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\\n40\\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\\n41\\n approximately 93% of ESRD patients in the United States were included in the data collected.\\n40\\n\\n\",\n \"The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\\n42\\n using the methods described by Calvert et al.\\n43\\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\\n43\\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\\n44\\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\\n45\\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\",\n \"Lynne E. Pinkerton conceived the study and wrote the first draft of the manuscript. Stephen Bertke conducted the analysis. Travis L. Kubale, James H. Yiin, and Robert D. Daniels acquired the data and assembled the cohort. Matthew M. Dahm led the assessment of the exposure surrogates. All authors participated in the interpretation and presentation of results, approved the final manuscript, and agreed to be accountable for all aspects of the work.\",\n \"John Meyer declares that he has no conflict of interest in the review and publication decision regarding this article.\",\n \"The study received approvals from the Institutional Review Boards of the National Institute for Occupational Safety and Health and the National Cancer Institute. Informed consent was waived for this records‐based study.\",\n \"The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Cohort description","Exposure assessment","Ascertainment of vital status and ESRD","Analysis","RESULTS","DISCUSSION","AUTHOR CONTRIBUTIONS","CONFLICTS OF INTEREST","DISCLOSURE BY AJIM EDITOR OF RECORD","ETHICS APPROVAL AND INFORMED CONSENT","DISCLAIMER","Supporting information"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Cohort description\",\n \"Exposure assessment\",\n \"Ascertainment of vital status and ESRD\",\n \"Analysis\",\n \"RESULTS\",\n \"DISCUSSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"CONFLICTS OF INTEREST\",\n \"DISCLOSURE BY AJIM EDITOR OF RECORD\",\n \"ETHICS APPROVAL AND INFORMED CONSENT\",\n \"DISCLAIMER\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Firefighters may be at increased risk of developing chronic kidney disease because they perform strenuous work in hot environments. Strenuous work with heat stress and dehydration may cause renal impairment from clinical or subclinical rhabdomyolysis, hyperuricemia and urate crystalluria, and hyperosmolality‐induced release of vasopressin and activation of the aldose reductase‐fructokinase pathway.\n1\n, \n2\n, \n3\n, \n4\n, \n5\n, \n6\n, \n7\n, \n8\n, \n9\n, \n10\n, \n11\n Increased concern about these potential heat‐related effects on kidney function has arisen due to an epidemic of chronic kidney disease among sugarcane workers in Central America.\n12\n Chronic kidney disease in these workers has often progressed to end‐stage renal disease (ESRD).\n3\n, \n13\n Chronic kidney disease has also been observed among other heat‐exposed workers and in other regions.\n4\n, \n12\n, \n14\n, \n15\n, \n16\n, \n17\n However, few data exist on the risk of kidney disease among firefighters.\nFirefighters are at risk of heat stress from exposure to heat from fires, ambient heat during hot weather, and metabolic heat generated by strenuous physical work.\n18\n, \n19\n The heavy, impermeable personal protective equipment structural firefighters wear adds to their metabolic workload and impairs evaporative cooling.\n19\n Increased core body temperature and heat‐related illnesses, including rhabdomyolysis and fatal heat stroke, have been documented among firefighters and trainees.\n20\n, \n21\n, \n22\n, \n23\n, \n24\n, \n25\n, \n26\n, \n27\n, \n28\n Firefighters are also at risk of dehydration,\n29\n which can contribute to kidney injury.\n30\n, \n31\n\n\nWe previously evaluated mortality and cancer incidence among a large National Institute for Occupational Safety and Health (NIOSH) cohort of US firefighters from the Chicago, Philadelphia, and San Francisco fire departments and found an increased risk of several cancers including kidney cancer compared to the US general population.\n32\n, \n33\n Mortality from chronic and unspecified nephritis and renal failure and other renal sclerosis, based on the underlying cause of death, was elevated among Philadelphia firefighters (standardized mortality ratio = 1.35; 95% confidence interval [CI] = 1.03–1.73) but not Chicago or San Francisco firefighters.\n33\n However, chronic kidney disease is usually not the underlying cause of death even when mentioned on the death certificate.\n34\n In this study, we evaluate ESRD incidence in the NIOSH firefighter cohort compared to the general population and relationships of ESRD with surrogates of exposure. Chicago and Philadelphia have humid continental climates, whereas San Francisco has a Mediterranean climate.\n35\n From 1981 to 2010, the average annual number of days the temperature reached 32.2°C was 17 in Chicago, 25 in Philadelphia, and 2 in San Francisco.\n36\n\n","[SUBTITLE] Cohort description [SUBSECTION] The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\nThe previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\n[SUBTITLE] Exposure assessment [SUBSECTION] Exposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\nExposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\n[SUBTITLE] Ascertainment of vital status and ESRD [SUBSECTION] Vital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n\nVital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n\n[SUBTITLE] Analysis [SUBSECTION] The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\nThe incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.","The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\n32\n, \n33\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.","Exposure surrogates were calculated for the restricted cohort as described previously.\n37\n, \n38\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\n38\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.","Vital status was ascertained through 2016 as described previously.\n32\n, \n33\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\n39\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\n40\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\n41\n approximately 93% of ESRD patients in the United States were included in the data collected.\n40\n\n","The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\n42\n using the methods described by Calvert et al.\n43\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\n43\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\n44\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\n45\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.","Table 1 shows the demographic characteristics of the full and restricted cohorts. Most members of the full cohort were male (96.3%) and White (78.6%). Over three‐fourths (78.1%) of the restricted cohort of male firefighters were also White. The mean employment duration for both cohorts was 21 years. At the end of the follow‐up, 38.2% of the full cohort and 25.3% of the restricted cohort were deceased.\nCharacteristics of the full and restricted cohorts\nAbbreviation: IQR, interquartile range.\nMale firefighters of a known race who were hired in 1950 or later and employed for at least 1 year.\nVital status at the end of follow‐up (December 31, 2014) or the date of diagnosis, whichever was earlier.\nCell size <11 or reporting the exact number allows a cell size <11 to be derived.\nAs shown in Table 2, the incidence of all ESRD was less than expected among the full (SIR = 0.79; 95% CI = 0.69–0.89; observed = 247) and restricted (SIR: 0.73; 95% CI = 0.63–0.85; observed = 177) cohorts. The median age at the time of onset was 69 years (interquartile range = 61–77) and 67 years (interquartile range: 59–73) among the full and restricted cohorts, respectively. SIRs for ESRD due to metabolic disease (full cohort: 2.84; 95% CI = 0.76–7.26; restricted cohort: 1.92; 95% CI = 0.22–6.94) and ESRD due to collagen vascular disease (full cohort: 1.18; 95% CI = 0.32–3.01; restricted cohort: 1.17; 95% CI = 0.24–3.42) were elevated, although the CIs for these ESRD types were wide and included one. The incidence of systemic ESRD and diabetic ESRD was less than expected among the full (systemic ESRD: SIR = 0.76; 95% CI = 0.65–0.88; diabetic ESRD: SIR = 0.66; 95% CI = 0.52–0.82) and restricted (systemic ESRD: SIR = 0.71; 95% CI = 059–0.84; diabetic ESRD: SIR = 0.63; 95% CI = 0.49–0.81) cohorts. The incidence of ESRD caused by glomerulonephritis was also less than expected among the full cohort (SIR = 0.61; 95% CI = 0.35–0.97). The SIR for ESRD caused by glomerulonephritis among the restricted cohort was 0.66 (95% CI = 0.36–1.10).\nSIRs for ESRD among the full and restricted cohorts, 1977–2014\nAbbreviations: AIDS, acquired immunodeficiency syndrome; CI, confidence interval; ESRD, end‐stage renal disease; NR, not reported (cell size <11 or reporting the cell size allows numbers <11 to be derived); Obs, observed; SIR, standardized incidence ratio.\nESRD due to AIDS, sickle cell disease, and hereditary nephropathy were omitted because no cases were observed.\nThe results of internal analyses among the restricted cohort using categorical models are shown in Table 3 (20‐year lag) and Supporting Information: Table S2 (10‐year lag). Results reported herein were obtained from the analyses with matching on employment duration, unless stated otherwise. HRs for all ESRD obtained from categorical models increased in the second and third quartiles of exposed days, but then decreased in the top quartile. HRs for systemic ESRD increased with each exposed‐day quartile in 20‐ and 10‐year lagged analyses. HRs for hypertensive ESRD also increased with the exposed‐day quartile in 20‐ and 10‐year lagged analyses with an HR of 4.51 (95% CI = 0.97–20.9) in the top quartile in 20‐year lagged analyses. HRs for diabetic ESRD increased with exposed‐day quartile in 10‐year lagged analyses only.\nESRD HRs by 20‐year lagged exposure surrogates obtained from categorical models\na\n\n\nAbbreviations: CFD, Chicago fire department; CI, confidence interval; dept, fire departments included in the analysis; ESRD, end‐stage renal disease; HR, hazard ratio; PFD, Philadelphia fire department; SFFD, San Francisco fire department.\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases with an approximately equal number of cases in each exposure quartile. The overall numbers of ESRD cases in the analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\nThe results of internal analyses among the restricted cohort using log‐linear models are shown in Table 4 (20‐year lag) and Supporting Information: Table S3 (10‐year lag). The HR for all ESRD was 1.11 (95% CI = 0.99–1.25) and 1.08 (95% CI = 0.98–1.18) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top exposure decile was excluded. In analogous linear models including person‐time in the top decile of exposure, the HR for all ESRD was 1.05 (95% CI = 0.96–1.16) and 1.07 (0.99–1.16) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively.\nESRD HRs by 20‐year lagged exposure surrogates obtained from log‐linear models\na\n\n\nAbbreviations: CI, confidence interval; ESRD, end‐stage renal disease; HR, hazard ratio.\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases. The overall numbers of ESRD cases in analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\nExposed days were available for Chicago, Philadelphia, and San Francisco firefighters; fire runs for Chicago and Philadelphia firefighters; and fire hours for Chicago firefighters.\n1000 exposed days, 1000 fire runs, or 1000 fire hours.\nSystemic ESRD HRs per 1000 exposed days were 1.05 (95% CI = 0.95–1.18) and 1.07 (95% CI = 0.98–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was included and 1.07 (95% CI = 0.95–1.23) and 1.05 (95% CI = 0.95–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was excluded. Hypertensive ESRD HRs per 1000 exposed days were 1.09 (95% CI = 0.93–1.29) and 1.10 (95% CI = 0.96–1.27) in 20‐ and 10‐year lagged analyses, respectively. When person‐time in the top exposure decile was excluded, the hypertensive ESRD HRs per 1000 exposed days were 1.13 (95% CI = 0.93–1.40) and 1.09 (95% CI = 0.94–1.29) in 20‐ and 10‐year lagged analyses, respectively. In 10‐year lagged analyses, the diabetic ESRD HR per 1000 exposed days was 1.03 (95% CI = 0.92–1.17) and 1.03 (95% CI = 0.90–1.19), including and excluding person‐time in the top exposure decile, respectively.","The incidence of all ESRD, systemic ESRD, and diabetic ESRD was less than expected based on rates of ESRD in the general population, which is consistent with a healthy worker effect. Diabetes mellitus is the most common cause of ESRD in the United States with over 45% of incident ESRD attributed to diabetes in 2016.\n41\n Thus, these findings may reflect, in part, differences in the prevalence and control of diabetes mellitus among the cohort compared to the general population. This is supported by the previously reported mortality from diabetes mellitus in the mortality cohort (SMR = 0.73; 95% CI = 0.64–0.83).\n33\n A lower prevalence of hyperglycemia among male US firefighters than the general population has also been reported.\n46\n\n\nIncreasing HRs were observed for systemic and hypertensive ESRD with exposed days in 20‐ and 10‐year lagged analyses and for diabetic ESRD with exposed days in 10‐year lagged analyses using categorical models. In log‐linear models, positive trends were observed for all ESRD, systemic ESRD, and hypertensive ESRD with exposed days, although the CIs included one. These associations with exposed days, but not fire runs and fire hours, may reflect the limitation of using fire runs and fire hours as surrogates for heat exposure. Firefighters are also exposed to occupational heat exposure and experience heat‐related illnesses during other activities such as training.\n21\n, \n27\n For the purposes of this study, we assumed that all three exposure surrogates would estimate occupational heat exposure in addition to exposure to combustion byproducts of fire. However, the exposure surrogates are likely crude measures of exposure in this study because we were unable to include additional information in the job exposure matrices regarding risk factors for heat exposure such as environmental air temperature and humidity. Other potential reasons for observing trends with exposed days but not fire runs or fire hours include the lack of data on fire runs and fire hours for all three fire departments and the potential impact of shift work on kidney function.\n47\n, \n48\n\n\nThe positive trend in hypertensive ESRD with exposed days, although not statistically significant, raises the possibility that the work of firefighters may exacerbate underlying hypertensive kidney disease. This trend could also potentially be due to an association between hypertension and other occupational exposures. Occupational exposures that may increase the risk of hypertension include particulate matter from fire smoke,\n49\n shift work,\n50\n work‐related stress,\n51\n noise,\n52\n, \n53\n and perfluoroalkyl substances.\n54\n, \n55\n In a recent study, 69% of firefighters had hypertension and the prevalence of hypertension was higher among male firefighters than in the general population.\n56\n\n\nIn the epidemic of chronic kidney disease in Central America, kidney disease rarely co‐occurred with diabetes or hypertension.\n57\n Biopsies from affected Central American workers showed interstitial nephritis.\n58\n A limitation of the current study was the inability to evaluate the association of ESRD due to interstitial nephritis with exposure surrogates due to the small number of observed cases.\nThe current study also does not provide information on the risk of chronic kidney disease that does not progress to ESRD. Other limitations include the lack of data on occupational heat exposure, surrogates of prolonged occupational heat exposure, and risk factors for ESRD such as hypertension,\n59\n diabetes mellitus,\n60\n and obesity.\n61\n In addition, cumulative exposure surrogates were negatively associated with leaving employment, which suggests that HRs obtained from analyses using g‐estimation instead of matching on employment duration to account for the healthy worker survivor effect may be lower. The findings are also not generalizable to wildland firefighters because of differences in personal protective equipment (heavy encapsulating turnout gear with self‐contained breathing apparatus vs. light, somewhat breathable fabrics and no respiratory protection), and the typical amount and duration of physical exertion (short bursts of very strenuous work for about 30 min a few times during a 24‐h shift vs. relatively constant strenuous work during a 12‐h shift each day during a 14‐day deployment).\nIn summary, the firefighters in this study were not at an increased risk of ESRD compared to the general population and statistically significant associations of ESRD with exposure surrogates were not observed. Only nonstatistically significant positive associations of all ESRD, systemic ESRD, and hypertensive ESRD with exposed days were observed. Associations of ESRD due to interstitial nephritis with surrogates of exposure could not be assessed due to the small number of observed cases. Additional research using more sensitive outcome measures to evaluate potential kidney effects among firefighters is warranted.","Lynne E. Pinkerton conceived the study and wrote the first draft of the manuscript. Stephen Bertke conducted the analysis. Travis L. Kubale, James H. Yiin, and Robert D. Daniels acquired the data and assembled the cohort. Matthew M. Dahm led the assessment of the exposure surrogates. All authors participated in the interpretation and presentation of results, approved the final manuscript, and agreed to be accountable for all aspects of the work.","Lynne E. Pinkerton currently works for Maximus, Inc., which holds contracts with CDC NIOSH for assistance, including the conduct of research studies. The remaining authors declare that there are no conflicts of interest.","John Meyer declares that he has no conflict of interest in the review and publication decision regarding this article.","The study received approvals from the Institutional Review Boards of the National Institute for Occupational Safety and Health and the National Cancer Institute. Informed consent was waived for this records‐based study.","The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention.","Supporting information.\nClick here for additional data file."],"string":"[\n \"Firefighters may be at increased risk of developing chronic kidney disease because they perform strenuous work in hot environments. Strenuous work with heat stress and dehydration may cause renal impairment from clinical or subclinical rhabdomyolysis, hyperuricemia and urate crystalluria, and hyperosmolality‐induced release of vasopressin and activation of the aldose reductase‐fructokinase pathway.\\n1\\n, \\n2\\n, \\n3\\n, \\n4\\n, \\n5\\n, \\n6\\n, \\n7\\n, \\n8\\n, \\n9\\n, \\n10\\n, \\n11\\n Increased concern about these potential heat‐related effects on kidney function has arisen due to an epidemic of chronic kidney disease among sugarcane workers in Central America.\\n12\\n Chronic kidney disease in these workers has often progressed to end‐stage renal disease (ESRD).\\n3\\n, \\n13\\n Chronic kidney disease has also been observed among other heat‐exposed workers and in other regions.\\n4\\n, \\n12\\n, \\n14\\n, \\n15\\n, \\n16\\n, \\n17\\n However, few data exist on the risk of kidney disease among firefighters.\\nFirefighters are at risk of heat stress from exposure to heat from fires, ambient heat during hot weather, and metabolic heat generated by strenuous physical work.\\n18\\n, \\n19\\n The heavy, impermeable personal protective equipment structural firefighters wear adds to their metabolic workload and impairs evaporative cooling.\\n19\\n Increased core body temperature and heat‐related illnesses, including rhabdomyolysis and fatal heat stroke, have been documented among firefighters and trainees.\\n20\\n, \\n21\\n, \\n22\\n, \\n23\\n, \\n24\\n, \\n25\\n, \\n26\\n, \\n27\\n, \\n28\\n Firefighters are also at risk of dehydration,\\n29\\n which can contribute to kidney injury.\\n30\\n, \\n31\\n\\n\\nWe previously evaluated mortality and cancer incidence among a large National Institute for Occupational Safety and Health (NIOSH) cohort of US firefighters from the Chicago, Philadelphia, and San Francisco fire departments and found an increased risk of several cancers including kidney cancer compared to the US general population.\\n32\\n, \\n33\\n Mortality from chronic and unspecified nephritis and renal failure and other renal sclerosis, based on the underlying cause of death, was elevated among Philadelphia firefighters (standardized mortality ratio = 1.35; 95% confidence interval [CI] = 1.03–1.73) but not Chicago or San Francisco firefighters.\\n33\\n However, chronic kidney disease is usually not the underlying cause of death even when mentioned on the death certificate.\\n34\\n In this study, we evaluate ESRD incidence in the NIOSH firefighter cohort compared to the general population and relationships of ESRD with surrogates of exposure. Chicago and Philadelphia have humid continental climates, whereas San Francisco has a Mediterranean climate.\\n35\\n From 1981 to 2010, the average annual number of days the temperature reached 32.2°C was 17 in Chicago, 25 in Philadelphia, and 2 in San Francisco.\\n36\\n\\n\",\n \"[SUBTITLE] Cohort description [SUBSECTION] The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\\n32\\n, \\n33\\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\\nThe previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\\n32\\n, \\n33\\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\\n[SUBTITLE] Exposure assessment [SUBSECTION] Exposure surrogates were calculated for the restricted cohort as described previously.\\n37\\n, \\n38\\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\\n38\\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\\nExposure surrogates were calculated for the restricted cohort as described previously.\\n37\\n, \\n38\\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\\n38\\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\\n[SUBTITLE] Ascertainment of vital status and ESRD [SUBSECTION] Vital status was ascertained through 2016 as described previously.\\n32\\n, \\n33\\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\\n39\\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\\n40\\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\\n41\\n approximately 93% of ESRD patients in the United States were included in the data collected.\\n40\\n\\n\\nVital status was ascertained through 2016 as described previously.\\n32\\n, \\n33\\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\\n39\\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\\n40\\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\\n41\\n approximately 93% of ESRD patients in the United States were included in the data collected.\\n40\\n\\n\\n[SUBTITLE] Analysis [SUBSECTION] The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\\n42\\n using the methods described by Calvert et al.\\n43\\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\\n43\\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\\n44\\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\\n45\\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\\nThe incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\\n42\\n using the methods described by Calvert et al.\\n43\\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\\n43\\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\\n44\\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\\n45\\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\",\n \"The previous mortality cohort included 29,992 career firefighters employed by the fire departments of Chicago (CFD), Philadelphia (PFD), and San Francisco (SFFD) for at least 1 day between the years 1950 and 2009, as described previously.\\n32\\n, \\n33\\n The current study cohort includes 26,469 members of the mortality cohort who were alive in 1977, when retrospective ascertainment of incident ESRD began, or later. Firefighters of unknown race (n = 727) were assumed White because 81% of firefighters of known race in the current cohort were White and 70% of firefighters of unknown race were hired before 1970 when minority hiring was lower. The cohort used in the exposure–response analysis, hereafter referred to as the restricted cohort, was further limited to 18,923 male firefighters of known race hired in 1950 or later for at least 1 year.\\nThe study received approvals from the Institutional Review Boards of NIOSH and the National Cancer Institute. Informed consent was waived for this records‐based study.\",\n \"Exposure surrogates were calculated for the restricted cohort as described previously.\\n37\\n, \\n38\\n These surrogates were originally developed to estimate exposure to combustion byproducts of fire for a cancer study,\\n38\\n but are used in this study as crude surrogates of occupational heat exposure. Briefly, detailed work histories through 2009 were linked with job exposure matrices based on the job, location, and fire‐fighting apparatus assignments. Data were available to calculate the number of exposed days (i.e., days worked in a job or location with potential exposure) for CFD, PFD, and SFFD firefighters; the number of fire runs for CFD and PFD firefighters; and the number of fire hours (i.e., the time spent at fires) for CFD firefighters.\",\n \"Vital status was ascertained through 2016 as described previously.\\n32\\n, \\n33\\n Cohort members with incident ESRD treated with dialysis or transplant from 1977 through 2014 were identified by linking the cohort with Medicare data on persons with ESRD by Social Security number (SSN), name, date of birth, and gender using the Fuzzy Lookup Transformation in Microsoft SQL Server Integration Services.\\n39\\n Inexact matches were manually reviewed in descending order of the overall similarity score until they were consistently judged to be false matches. All inexact matches with an exact match on the Social Security number were also manually reviewed.\\nMedicare data on persons with ESRD is maintained by the Centers for Medicare & Medicaid Services (CMS) and includes individuals who received Medicare‐covered renal replacement therapy (dialysis or transplant) in 1977 or later. Since 1973, most persons with ESRD have been entitled to Medicare regardless of age.\\n40\\n Between the earliest years, when reporting was incomplete, and 1995, when reporting of all ESRD patients regardless of Medicare eligibility was mandated,\\n41\\n approximately 93% of ESRD patients in the United States were included in the data collected.\\n40\\n\\n\",\n \"The incidence of ESRD among the overall cohort was compared to that of the US general population in R version 4.1.3\\n42\\n using the methods described by Calvert et al.\\n43\\n Since the Medicare data on ESRD are incomplete before 1977, cohort members who died before 1977 were excluded from the analyses. Person‐years‐at‐risk (PYAR) were accumulated from the date the cohort criteria were met or January 1, 1977, whichever was later, to the first service date for ESRD, the date of death, the date last observed, or the ending date of the study (December 31, 2014), whichever was earliest. The first service date for ESRD is generally the date on which renal replacement therapy began and was used as a surrogate for the date of onset for ESRD. PYAR were stratified by gender, race (White, other), and 5‐year intervals of age and calendar time and then multiplied by the appropriate US ESRD incidence rates to calculate the expected number of cases for each stratum. The US incidence rates were created from CMS Medicare data and US census data as described elsewhere.\\n43\\n The expected number of ESRD cases in each stratum was summed to obtain the total expected number. The standardized incidence ratio (SIR) was calculated as the ratio of the observed to the expected number of treated ESRD cases. SIRs were also calculated for types of ESRD and for the restricted cohort. Ninety‐five percent CIs were computed assuming a Poisson distribution for observed ESRD cases. Types of ESRD and the corresponding International Classification of Diseases, 9th Revision, Clinical Modification codes for the primary cause of renal failure in the CMS data are listed in Supporting Information: Table S1.\\nExposure–response associations within the restricted cohort were examined with Cox proportional hazards regression using the SAS PHREG procedure.\\n44\\n Risk sets comprised all restricted cohort members at risk as of the attained age of the case who also matched the case on race (White, other), birth date (within 5 years), and fire department. Restricted cohort members were followed from the completion of the 1‐year eligibility period or January 1, 1977, whichever was latest. The cumulative exposure metrics were lagged 10 and 20 years because chronic kidney failure generally develops over many years. In categorical models, the cut‐points were selected to obtain an approximately equal number of cases in each exposure quartile. Hazard ratios (HRs) were estimated from the maximum partial likelihood ratio test and two‐sided 95% CIs were based on the profile likelihood. The exposure surrogates were also analyzed as continuous time‐dependent variables in log‐linear models where the HR increases exponentially with exposure or, equivalently, the log of the HR increases linearly with exposure. Because the exposure surrogates were generally right‐skewed and exposure–response relationships in occupational studies are often attenuated at high exposure levels,\\n45\\n log‐linear models were also evaluated excluding person‐time in the top decile of exposure. The healthy worker survivor effect was addressed by repeating the analyses also matching on employment duration (<10, 10–<20, 20–<30, and 30+ years). Modeling analyses were restricted to analyses with 30 or more observed cases.\",\n \"Table 1 shows the demographic characteristics of the full and restricted cohorts. Most members of the full cohort were male (96.3%) and White (78.6%). Over three‐fourths (78.1%) of the restricted cohort of male firefighters were also White. The mean employment duration for both cohorts was 21 years. At the end of the follow‐up, 38.2% of the full cohort and 25.3% of the restricted cohort were deceased.\\nCharacteristics of the full and restricted cohorts\\nAbbreviation: IQR, interquartile range.\\nMale firefighters of a known race who were hired in 1950 or later and employed for at least 1 year.\\nVital status at the end of follow‐up (December 31, 2014) or the date of diagnosis, whichever was earlier.\\nCell size <11 or reporting the exact number allows a cell size <11 to be derived.\\nAs shown in Table 2, the incidence of all ESRD was less than expected among the full (SIR = 0.79; 95% CI = 0.69–0.89; observed = 247) and restricted (SIR: 0.73; 95% CI = 0.63–0.85; observed = 177) cohorts. The median age at the time of onset was 69 years (interquartile range = 61–77) and 67 years (interquartile range: 59–73) among the full and restricted cohorts, respectively. SIRs for ESRD due to metabolic disease (full cohort: 2.84; 95% CI = 0.76–7.26; restricted cohort: 1.92; 95% CI = 0.22–6.94) and ESRD due to collagen vascular disease (full cohort: 1.18; 95% CI = 0.32–3.01; restricted cohort: 1.17; 95% CI = 0.24–3.42) were elevated, although the CIs for these ESRD types were wide and included one. The incidence of systemic ESRD and diabetic ESRD was less than expected among the full (systemic ESRD: SIR = 0.76; 95% CI = 0.65–0.88; diabetic ESRD: SIR = 0.66; 95% CI = 0.52–0.82) and restricted (systemic ESRD: SIR = 0.71; 95% CI = 059–0.84; diabetic ESRD: SIR = 0.63; 95% CI = 0.49–0.81) cohorts. The incidence of ESRD caused by glomerulonephritis was also less than expected among the full cohort (SIR = 0.61; 95% CI = 0.35–0.97). The SIR for ESRD caused by glomerulonephritis among the restricted cohort was 0.66 (95% CI = 0.36–1.10).\\nSIRs for ESRD among the full and restricted cohorts, 1977–2014\\nAbbreviations: AIDS, acquired immunodeficiency syndrome; CI, confidence interval; ESRD, end‐stage renal disease; NR, not reported (cell size <11 or reporting the cell size allows numbers <11 to be derived); Obs, observed; SIR, standardized incidence ratio.\\nESRD due to AIDS, sickle cell disease, and hereditary nephropathy were omitted because no cases were observed.\\nThe results of internal analyses among the restricted cohort using categorical models are shown in Table 3 (20‐year lag) and Supporting Information: Table S2 (10‐year lag). Results reported herein were obtained from the analyses with matching on employment duration, unless stated otherwise. HRs for all ESRD obtained from categorical models increased in the second and third quartiles of exposed days, but then decreased in the top quartile. HRs for systemic ESRD increased with each exposed‐day quartile in 20‐ and 10‐year lagged analyses. HRs for hypertensive ESRD also increased with the exposed‐day quartile in 20‐ and 10‐year lagged analyses with an HR of 4.51 (95% CI = 0.97–20.9) in the top quartile in 20‐year lagged analyses. HRs for diabetic ESRD increased with exposed‐day quartile in 10‐year lagged analyses only.\\nESRD HRs by 20‐year lagged exposure surrogates obtained from categorical models\\na\\n\\n\\nAbbreviations: CFD, Chicago fire department; CI, confidence interval; dept, fire departments included in the analysis; ESRD, end‐stage renal disease; HR, hazard ratio; PFD, Philadelphia fire department; SFFD, San Francisco fire department.\\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases with an approximately equal number of cases in each exposure quartile. The overall numbers of ESRD cases in the analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\\nThe results of internal analyses among the restricted cohort using log‐linear models are shown in Table 4 (20‐year lag) and Supporting Information: Table S3 (10‐year lag). The HR for all ESRD was 1.11 (95% CI = 0.99–1.25) and 1.08 (95% CI = 0.98–1.18) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top exposure decile was excluded. In analogous linear models including person‐time in the top decile of exposure, the HR for all ESRD was 1.05 (95% CI = 0.96–1.16) and 1.07 (0.99–1.16) per 1000 exposed days in 20‐ and 10‐year lagged analyses, respectively.\\nESRD HRs by 20‐year lagged exposure surrogates obtained from log‐linear models\\na\\n\\n\\nAbbreviations: CI, confidence interval; ESRD, end‐stage renal disease; HR, hazard ratio.\\nAll models were adjusted for age (time scale) and for race (White/non‐White), date of birth (within 5 years), and fire department by matching. Some models were also adjusted for employment duration (<10, 10–<20, 20–<30, and 30+ years) by matching.\\nAll modeling analyses were conducted among the restricted cohort and included 30 or more observed ESRD cases. The overall numbers of ESRD cases in analyses of exposed days were 177 for all ESRD, 132 for systemic ESRD, 63 for diabetic ESRD, and 53 for hypertensive ESRD.\\nExposed days were available for Chicago, Philadelphia, and San Francisco firefighters; fire runs for Chicago and Philadelphia firefighters; and fire hours for Chicago firefighters.\\n1000 exposed days, 1000 fire runs, or 1000 fire hours.\\nSystemic ESRD HRs per 1000 exposed days were 1.05 (95% CI = 0.95–1.18) and 1.07 (95% CI = 0.98–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was included and 1.07 (95% CI = 0.95–1.23) and 1.05 (95% CI = 0.95–1.17) in 20‐ and 10‐year lagged analyses, respectively, when person‐time in the top decile of exposure was excluded. Hypertensive ESRD HRs per 1000 exposed days were 1.09 (95% CI = 0.93–1.29) and 1.10 (95% CI = 0.96–1.27) in 20‐ and 10‐year lagged analyses, respectively. When person‐time in the top exposure decile was excluded, the hypertensive ESRD HRs per 1000 exposed days were 1.13 (95% CI = 0.93–1.40) and 1.09 (95% CI = 0.94–1.29) in 20‐ and 10‐year lagged analyses, respectively. In 10‐year lagged analyses, the diabetic ESRD HR per 1000 exposed days was 1.03 (95% CI = 0.92–1.17) and 1.03 (95% CI = 0.90–1.19), including and excluding person‐time in the top exposure decile, respectively.\",\n \"The incidence of all ESRD, systemic ESRD, and diabetic ESRD was less than expected based on rates of ESRD in the general population, which is consistent with a healthy worker effect. Diabetes mellitus is the most common cause of ESRD in the United States with over 45% of incident ESRD attributed to diabetes in 2016.\\n41\\n Thus, these findings may reflect, in part, differences in the prevalence and control of diabetes mellitus among the cohort compared to the general population. This is supported by the previously reported mortality from diabetes mellitus in the mortality cohort (SMR = 0.73; 95% CI = 0.64–0.83).\\n33\\n A lower prevalence of hyperglycemia among male US firefighters than the general population has also been reported.\\n46\\n\\n\\nIncreasing HRs were observed for systemic and hypertensive ESRD with exposed days in 20‐ and 10‐year lagged analyses and for diabetic ESRD with exposed days in 10‐year lagged analyses using categorical models. In log‐linear models, positive trends were observed for all ESRD, systemic ESRD, and hypertensive ESRD with exposed days, although the CIs included one. These associations with exposed days, but not fire runs and fire hours, may reflect the limitation of using fire runs and fire hours as surrogates for heat exposure. Firefighters are also exposed to occupational heat exposure and experience heat‐related illnesses during other activities such as training.\\n21\\n, \\n27\\n For the purposes of this study, we assumed that all three exposure surrogates would estimate occupational heat exposure in addition to exposure to combustion byproducts of fire. However, the exposure surrogates are likely crude measures of exposure in this study because we were unable to include additional information in the job exposure matrices regarding risk factors for heat exposure such as environmental air temperature and humidity. Other potential reasons for observing trends with exposed days but not fire runs or fire hours include the lack of data on fire runs and fire hours for all three fire departments and the potential impact of shift work on kidney function.\\n47\\n, \\n48\\n\\n\\nThe positive trend in hypertensive ESRD with exposed days, although not statistically significant, raises the possibility that the work of firefighters may exacerbate underlying hypertensive kidney disease. This trend could also potentially be due to an association between hypertension and other occupational exposures. Occupational exposures that may increase the risk of hypertension include particulate matter from fire smoke,\\n49\\n shift work,\\n50\\n work‐related stress,\\n51\\n noise,\\n52\\n, \\n53\\n and perfluoroalkyl substances.\\n54\\n, \\n55\\n In a recent study, 69% of firefighters had hypertension and the prevalence of hypertension was higher among male firefighters than in the general population.\\n56\\n\\n\\nIn the epidemic of chronic kidney disease in Central America, kidney disease rarely co‐occurred with diabetes or hypertension.\\n57\\n Biopsies from affected Central American workers showed interstitial nephritis.\\n58\\n A limitation of the current study was the inability to evaluate the association of ESRD due to interstitial nephritis with exposure surrogates due to the small number of observed cases.\\nThe current study also does not provide information on the risk of chronic kidney disease that does not progress to ESRD. Other limitations include the lack of data on occupational heat exposure, surrogates of prolonged occupational heat exposure, and risk factors for ESRD such as hypertension,\\n59\\n diabetes mellitus,\\n60\\n and obesity.\\n61\\n In addition, cumulative exposure surrogates were negatively associated with leaving employment, which suggests that HRs obtained from analyses using g‐estimation instead of matching on employment duration to account for the healthy worker survivor effect may be lower. The findings are also not generalizable to wildland firefighters because of differences in personal protective equipment (heavy encapsulating turnout gear with self‐contained breathing apparatus vs. light, somewhat breathable fabrics and no respiratory protection), and the typical amount and duration of physical exertion (short bursts of very strenuous work for about 30 min a few times during a 24‐h shift vs. relatively constant strenuous work during a 12‐h shift each day during a 14‐day deployment).\\nIn summary, the firefighters in this study were not at an increased risk of ESRD compared to the general population and statistically significant associations of ESRD with exposure surrogates were not observed. Only nonstatistically significant positive associations of all ESRD, systemic ESRD, and hypertensive ESRD with exposed days were observed. Associations of ESRD due to interstitial nephritis with surrogates of exposure could not be assessed due to the small number of observed cases. Additional research using more sensitive outcome measures to evaluate potential kidney effects among firefighters is warranted.\",\n \"Lynne E. Pinkerton conceived the study and wrote the first draft of the manuscript. Stephen Bertke conducted the analysis. Travis L. Kubale, James H. Yiin, and Robert D. Daniels acquired the data and assembled the cohort. Matthew M. Dahm led the assessment of the exposure surrogates. All authors participated in the interpretation and presentation of results, approved the final manuscript, and agreed to be accountable for all aspects of the work.\",\n \"Lynne E. Pinkerton currently works for Maximus, Inc., which holds contracts with CDC NIOSH for assistance, including the conduct of research studies. The remaining authors declare that there are no conflicts of interest.\",\n \"John Meyer declares that he has no conflict of interest in the review and publication decision regarding this article.\",\n \"The study received approvals from the Institutional Review Boards of the National Institute for Occupational Safety and Health and the National Cancer Institute. Informed consent was waived for this records‐based study.\",\n \"The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention.\",\n \"Supporting information.\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,null,null,null,"results","discussion",null,"COI-statement",null,null,null,"supplementary-material"],"string":"[\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n \"results\",\n \"discussion\",\n null,\n \"COI-statement\",\n null,\n null,\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["end‐stage renal disease","firefighters","longitudinal studies"],"string":"[\n \"end‐stage renal disease\",\n \"firefighters\",\n \"longitudinal studies\"\n]"}}},{"rowIdx":382,"cells":{"title":{"kind":"string","value":"Analysis of phenotype and gene mutation in three pedigrees with inherited antithrombin deficiency."},"pmid":{"kind":"string","value":"36268972"},"background_abstract":{"kind":"string","value":"Inherited AT deficiency is an autosomal-dominant thrombophilic disorder usually caused by various SERPINC1 defects associated with a high risk of recurrent venous thromboembolism. In this article, the phenotype, gene mutation, and molecular pathogenic mechanisms were determined in three pedigrees with inherited AT deficiency."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Coagulation indices were examined on STAGO STA-R-MAX analyzer. The AT:Ag was analyzed by ELISA. All exons and flanking sequences of SERPINC1 were amplified by PCR. AT wild type and three mutant expression plasmids were constructed and then transfected into HEK293FT cells. The expression level of AT protein was analyzed by ELISA and Western blot."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The AT:A and AT:Ag of probands 1 and 3 were decreased to 49% and 52 mg/dL, 38% and 44 mg/dL, respectively. The AT:A of proband 2 was decreased to 32%. The SERPINC1 gene analysis indicated that there was a p.Ile421Thr in proband 1, a p.Leu417Gln in proband 2, and a p.Met252Thr in proband 3, respectively. The AT mRNA expression level of the three mutants was not significantly different from AT-WT by qRT-PCR. The results of ELISA and Western blot tests showed that the AT-M252T and AT-I421T mutants had a higher AT expression than the AT wild type (AT-WT), and the AT protein expression of AT-L417Q mutants had no significant difference compared with AT-WT in the cell lysate. The AT expression levels of AT-M252T and AT-I421T mutants were lower than that of AT-WT, and there was no significant difference between AT-L417Q mutant and AT-WT in the supernatant."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The p.I421T and p.M252T mutations affected the secretion of AT protein leading to type I AT deficiency of probands 1 and 3. The p.Leu417Gln mutation was responsible for the impaired or ineffective activity AT protein in proband 2 and caused type II AT deficiency."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Pedigree","Antithrombin III Deficiency","Phenotype","Mutation","Antithrombins"],"string":"[\n \"Humans\",\n \"Pedigree\",\n \"Antithrombin III Deficiency\",\n \"Phenotype\",\n \"Mutation\",\n \"Antithrombins\"\n]"},"pmcid":{"kind":"string","value":"9701880"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Phenotype and genotype [SUBSECTION] In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\nIn pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\n[SUBTITLE] Bioformatics [SUBSECTION] The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\nThe bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\n[SUBTITLE] Transfection and expression [SUBSECTION] The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\nThe AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results."},"conclusions_title":{"kind":"string","value":"CONCLUSION"},"conclusions_text":{"kind":"string","value":"We reported three heterozygous missense mutations in the SERPINC1 gene, namely c.1358 T > C (p.Ile421Thr), c.1346 T > A (p.Leu417Gln), and c.851 T > C (p.Met252Thr). The pathogenic mechanisms of the three mutations were preliminarily explored through bioinformatics, protein modeling, and in vitro expression analysis, which enriched the AT genetic mutation database."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Patients","Plasma AT assays","\nAT gene analysis","Bioformatics analysis","Construction, transfection, and purification","Phenotype and genotype","Bioformatics","Transfection and expression","AUTHOR CONTRIBUTIONS","ETHICS STATEMENT"],"string":"[\n \"INTRODUCTION\",\n \"Patients\",\n \"Plasma AT assays\",\n \"\\nAT gene analysis\",\n \"Bioformatics analysis\",\n \"Construction, transfection, and purification\",\n \"Phenotype and genotype\",\n \"Bioformatics\",\n \"Transfection and expression\",\n \"AUTHOR CONTRIBUTIONS\",\n \"ETHICS STATEMENT\"\n]"},"other_sections_texts":{"kind":"list like","value":["Antithrombin (AT) is the major serine protease inhibitor (SERPIN) in plasma that regulates the proteolytic activities of coagulation proteases of both intrinsic and extrinsic pathways. AT is synthesized predominantly by hepatocytes, has a half‐life of approximately 2.4 days and a molecular weight of 58 kDa, and contains 432 amino acids.\n1\n AT is the major inhibitor of thrombin, activated coagulation factor IX (FIXa), and activated coagulation factor Х (FXa) in plasma, and is also able to inactivate the other serine proteases of the intrinsic coagulation pathway, activated factors XI and XII (FXIa and FXIIa).\n2\n In the physiologic state, AT circulates in a relatively inefficient inhibitor form. However, the anticoagulant effect of AT is greatly increased (at least 1000‐fold) by the presence of heparin and other heparin‐like glycosaminoglycans. This effect is the basis for the use of heparin and low‐molecular‐weight heparins (LMWH) as anticoagulants in patients with venous thromboembolism (VTE).\n3\n\n\nThe AT gene (SERPINC1) which encodes antithrombin is mapped on chromosome 1q23‐25 and comprises 7 exons encompassing 13.5 kb of genomic DNA.\n4\n The 1392 bp mRNA codes for a 464 amino acid polypeptide chain that undergoes several posttranslational modifications including cleavage of the signal peptide (32 amino acids) and N‐glycosylation.\n5\n, \n6\n Inherited AT deficiency is an autosomal‐dominant disorder caused by a defect in the SERPINC1 gene. It was first described in 1965 as one of the important risk factors for hemostatic diseases characterized by hereditary and recurrent thrombosis. In healthy people, the prevalence of inherited AT deficiency is around 0.02%–0.25%, but the incidence in patients with venous thrombosis is increased around 1%–2%.\n7\n, \n8\n, \n9\n Inherited AT deficiency is divided into type І with a reduction in the antigen levels and functional activity and type II deficiency with reduced antithrombin activity but almost normal antigen level.\n10\n, \n11\n\n\nMost cases with type I defects are heterozygous; homozygous patients have a higher risk of severe venous thromboembolism (VTE) in childhood and may even be fatal in utero.\n9\n For type II AT deficiency, the level of AT in the patient's plasma is not significantly affected; therefore, we speculate that type II has a slight thrombotic effect. Related studies have shown that patients with type II deficiency exhibit extensive and heterogeneous clinical phenotype.\n12\n\n\nInherited AT deficiency is widely considered to be one of the causes of thrombotic diseases. However, there are a few large cohort studies reported. More than 400 different mutations have been registered in the AT Mutation Database (The Human Gene Mutation Database, http://www.hgmd.cf.ac.uk/ac/all.php). Most are missense/nonsense mutations, the remaining being small insertion/deletion, large insertion/deletion, shear mutation, and complex mutation. The SERPINC1 gene mutations are heterogeneous in the population, and there are not many race‐specific mutations.\n13\n, \n14\n, \n15\n\n\nInherited AT deficiency is the main genetic risk factor for deep vein thrombosis (DVT) patients in China.\n16\n Sporadic cases with SERPINC1 gene mutations were reported in the population; however, no mutation hotspots were recognized. In this article, we presented three Chinese pedigrees with inherited AT deficiency, the phenotype, gene mutations, and expression of mutation sites in vitro. The molecular bases of the three pedigrees were analyzed.","Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\nThree pedigrees of inherited AT deficiency\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.","Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.","Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\n17\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.","The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).","The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.","In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively","The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.","The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.","MSW and ML designed experiments, analyzed data, wrote the article, and supervised the project. SQL and STJ performed molecular modeling. SQL and YHJ designed experiments, conducted genetic analysis, and coagulation assays. All authors approved the final version of the article.","Our study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China)."],"string":"[\n \"Antithrombin (AT) is the major serine protease inhibitor (SERPIN) in plasma that regulates the proteolytic activities of coagulation proteases of both intrinsic and extrinsic pathways. AT is synthesized predominantly by hepatocytes, has a half‐life of approximately 2.4 days and a molecular weight of 58 kDa, and contains 432 amino acids.\\n1\\n AT is the major inhibitor of thrombin, activated coagulation factor IX (FIXa), and activated coagulation factor Х (FXa) in plasma, and is also able to inactivate the other serine proteases of the intrinsic coagulation pathway, activated factors XI and XII (FXIa and FXIIa).\\n2\\n In the physiologic state, AT circulates in a relatively inefficient inhibitor form. However, the anticoagulant effect of AT is greatly increased (at least 1000‐fold) by the presence of heparin and other heparin‐like glycosaminoglycans. This effect is the basis for the use of heparin and low‐molecular‐weight heparins (LMWH) as anticoagulants in patients with venous thromboembolism (VTE).\\n3\\n\\n\\nThe AT gene (SERPINC1) which encodes antithrombin is mapped on chromosome 1q23‐25 and comprises 7 exons encompassing 13.5 kb of genomic DNA.\\n4\\n The 1392 bp mRNA codes for a 464 amino acid polypeptide chain that undergoes several posttranslational modifications including cleavage of the signal peptide (32 amino acids) and N‐glycosylation.\\n5\\n, \\n6\\n Inherited AT deficiency is an autosomal‐dominant disorder caused by a defect in the SERPINC1 gene. It was first described in 1965 as one of the important risk factors for hemostatic diseases characterized by hereditary and recurrent thrombosis. In healthy people, the prevalence of inherited AT deficiency is around 0.02%–0.25%, but the incidence in patients with venous thrombosis is increased around 1%–2%.\\n7\\n, \\n8\\n, \\n9\\n Inherited AT deficiency is divided into type І with a reduction in the antigen levels and functional activity and type II deficiency with reduced antithrombin activity but almost normal antigen level.\\n10\\n, \\n11\\n\\n\\nMost cases with type I defects are heterozygous; homozygous patients have a higher risk of severe venous thromboembolism (VTE) in childhood and may even be fatal in utero.\\n9\\n For type II AT deficiency, the level of AT in the patient's plasma is not significantly affected; therefore, we speculate that type II has a slight thrombotic effect. Related studies have shown that patients with type II deficiency exhibit extensive and heterogeneous clinical phenotype.\\n12\\n\\n\\nInherited AT deficiency is widely considered to be one of the causes of thrombotic diseases. However, there are a few large cohort studies reported. More than 400 different mutations have been registered in the AT Mutation Database (The Human Gene Mutation Database, http://www.hgmd.cf.ac.uk/ac/all.php). Most are missense/nonsense mutations, the remaining being small insertion/deletion, large insertion/deletion, shear mutation, and complex mutation. The SERPINC1 gene mutations are heterogeneous in the population, and there are not many race‐specific mutations.\\n13\\n, \\n14\\n, \\n15\\n\\n\\nInherited AT deficiency is the main genetic risk factor for deep vein thrombosis (DVT) patients in China.\\n16\\n Sporadic cases with SERPINC1 gene mutations were reported in the population; however, no mutation hotspots were recognized. In this article, we presented three Chinese pedigrees with inherited AT deficiency, the phenotype, gene mutations, and expression of mutation sites in vitro. The molecular bases of the three pedigrees were analyzed.\",\n \"Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\\nThree pedigrees of inherited AT deficiency\\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\",\n \"Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\",\n \"Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\\n17\\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\",\n \"The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\",\n \"The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.\",\n \"In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\\nPhenotype results of three pedigrees with inherited AT deficiency\\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\",\n \"The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\\n18\\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\\nBioinformatics prediction results of mutations\\n\\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\\nConservation analysis of homologous species of Met252, Leu417, and Ile421\\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\",\n \"The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\",\n \"MSW and ML designed experiments, analyzed data, wrote the article, and supervised the project. SQL and STJ performed molecular modeling. SQL and YHJ designed experiments, conducted genetic analysis, and coagulation assays. All authors approved the final version of the article.\",\n \"Our study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China).\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","MATERIALS AND METHODS","Patients","Plasma AT assays","\nAT gene analysis","Bioformatics analysis","Construction, transfection, and purification","RESULTS","Phenotype and genotype","Bioformatics","Transfection and expression","DISCUSSION","CONCLUSION","AUTHOR CONTRIBUTIONS","CONFLICT OF INTEREST","ETHICS STATEMENT","Supporting information"],"string":"[\n \"INTRODUCTION\",\n \"MATERIALS AND METHODS\",\n \"Patients\",\n \"Plasma AT assays\",\n \"\\nAT gene analysis\",\n \"Bioformatics analysis\",\n \"Construction, transfection, and purification\",\n \"RESULTS\",\n \"Phenotype and genotype\",\n \"Bioformatics\",\n \"Transfection and expression\",\n \"DISCUSSION\",\n \"CONCLUSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"CONFLICT OF INTEREST\",\n \"ETHICS STATEMENT\",\n \"Supporting information\"\n]"},"all_sections_texts":{"kind":"list like","value":["Antithrombin (AT) is the major serine protease inhibitor (SERPIN) in plasma that regulates the proteolytic activities of coagulation proteases of both intrinsic and extrinsic pathways. AT is synthesized predominantly by hepatocytes, has a half‐life of approximately 2.4 days and a molecular weight of 58 kDa, and contains 432 amino acids.\n1\n AT is the major inhibitor of thrombin, activated coagulation factor IX (FIXa), and activated coagulation factor Х (FXa) in plasma, and is also able to inactivate the other serine proteases of the intrinsic coagulation pathway, activated factors XI and XII (FXIa and FXIIa).\n2\n In the physiologic state, AT circulates in a relatively inefficient inhibitor form. However, the anticoagulant effect of AT is greatly increased (at least 1000‐fold) by the presence of heparin and other heparin‐like glycosaminoglycans. This effect is the basis for the use of heparin and low‐molecular‐weight heparins (LMWH) as anticoagulants in patients with venous thromboembolism (VTE).\n3\n\n\nThe AT gene (SERPINC1) which encodes antithrombin is mapped on chromosome 1q23‐25 and comprises 7 exons encompassing 13.5 kb of genomic DNA.\n4\n The 1392 bp mRNA codes for a 464 amino acid polypeptide chain that undergoes several posttranslational modifications including cleavage of the signal peptide (32 amino acids) and N‐glycosylation.\n5\n, \n6\n Inherited AT deficiency is an autosomal‐dominant disorder caused by a defect in the SERPINC1 gene. It was first described in 1965 as one of the important risk factors for hemostatic diseases characterized by hereditary and recurrent thrombosis. In healthy people, the prevalence of inherited AT deficiency is around 0.02%–0.25%, but the incidence in patients with venous thrombosis is increased around 1%–2%.\n7\n, \n8\n, \n9\n Inherited AT deficiency is divided into type І with a reduction in the antigen levels and functional activity and type II deficiency with reduced antithrombin activity but almost normal antigen level.\n10\n, \n11\n\n\nMost cases with type I defects are heterozygous; homozygous patients have a higher risk of severe venous thromboembolism (VTE) in childhood and may even be fatal in utero.\n9\n For type II AT deficiency, the level of AT in the patient's plasma is not significantly affected; therefore, we speculate that type II has a slight thrombotic effect. Related studies have shown that patients with type II deficiency exhibit extensive and heterogeneous clinical phenotype.\n12\n\n\nInherited AT deficiency is widely considered to be one of the causes of thrombotic diseases. However, there are a few large cohort studies reported. More than 400 different mutations have been registered in the AT Mutation Database (The Human Gene Mutation Database, http://www.hgmd.cf.ac.uk/ac/all.php). Most are missense/nonsense mutations, the remaining being small insertion/deletion, large insertion/deletion, shear mutation, and complex mutation. The SERPINC1 gene mutations are heterogeneous in the population, and there are not many race‐specific mutations.\n13\n, \n14\n, \n15\n\n\nInherited AT deficiency is the main genetic risk factor for deep vein thrombosis (DVT) patients in China.\n16\n Sporadic cases with SERPINC1 gene mutations were reported in the population; however, no mutation hotspots were recognized. In this article, we presented three Chinese pedigrees with inherited AT deficiency, the phenotype, gene mutations, and expression of mutation sites in vitro. The molecular bases of the three pedigrees were analyzed.","[SUBTITLE] Patients [SUBSECTION] Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\nThree pedigrees of inherited AT deficiency\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\nProband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\nThree pedigrees of inherited AT deficiency\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\n[SUBTITLE] Plasma AT assays [SUBSECTION] Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\nPeripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\n[SUBTITLE] \nAT gene analysis [SUBSECTION] Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\n17\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\nPeripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\n17\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\n[SUBTITLE] Bioformatics analysis [SUBSECTION] The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\nThe bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\n[SUBTITLE] Construction, transfection, and purification [SUBSECTION] The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.\nThe vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.","Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\nThree pedigrees of inherited AT deficiency\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.","Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.","Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\n17\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.","The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).","The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.","[SUBTITLE] Phenotype and genotype [SUBSECTION] In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\nIn pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\n[SUBTITLE] Bioformatics [SUBSECTION] The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\nThe bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\n[SUBTITLE] Transfection and expression [SUBSECTION] The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\nThe AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.","In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\nPhenotype results of three pedigrees with inherited AT deficiency\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively","The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\n18\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\nBioinformatics prediction results of mutations\n\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\nConservation analysis of homologous species of Met252, Leu417, and Ile421\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.","The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.","Thrombophilia refers to a high thromboembolic tendency caused by certain risk factors, including hereditary and acquired factors. Genetic factors mainly include congenital defects of anticoagulant proteins such as AT, PC, and PS, mutations of prothrombin G20210A, and coagulation factor V Leiden.\n19\n Acquired risk factors include, but are not limited to, oral contraceptives, trauma, immobilization, and surgery, and other factors.\n20\n In this article, we presented 3 families with inherited AT deficiency and found 3 heterozygous missense variants, namely c.1358 T > C (p.Ile421Thr), c.1346 T > A (p.Leu417Gln), and c.851 T > C (p.Met252Thr). However, these three variants were not found in 100 healthy controls, thus excluding genetic polymorphisms.\nProband 1 had p.Ile421Thr mutation and was diagnosed as a type I Inherited AT deficiency. The results of conservation analysis showed that the Ile421 site was highly conserved in homologous species, suggesting that this site plays an important role in the structure or function of AT protein. When le421 was replaced by Thr421, intermolecular hydrophobic force and hydrogen bonds changed, and the protein structure might have been affected, which was in accordance with the results of bioinformatics analysis. The p.Ile421Thr mutation has only been reported since 1994 by overseas researchers.\n20\n, \n21\n The overseas case presented recurrent venous thrombosis with a family history of thrombosis, and developed DVT and PE after the first pregnancy at the age of 27 years, with similar symptoms after the second and third pregnancies. Similar to our study, proband 1 was 27 years old and had venous thrombosis during puerperium, indicating that the p.Ile421Thr mutation has a higher risk of thrombosis, and when combined with acquired risk factors such as pregnancy and surgery, the p.Ile421Thr mutation has early onset characteristics. The p.Leu417Gln mutation was found in all four members in the second family (proband 2, his son, daughter, and granddaughter) who carried this heterozygous mutation; the details of which had been reported by our group.\n18\n All members in the third family including the proband 3, her mother, and uncle carried the p.Met252Thr heterozygous mutation. Western blot results showed that proband 3's plasma AT level was lower than the normal controls, in accordance with the phenotypic results, and diagnosed as a type I hereditary AT deficiency.\n22\n The bioinformatics software predicted that it would be pathogenic and that the protein structure may be affected. The hydrophobic interaction force and the number of hydrogen bonds between Thr252 and surrounding amino acids had been changed, which had an influence on the structural stability. In family 3, all three members carrying the heterozygous mutation of p.Met252Thr had a history of venous thrombosis; the wild‐type members were asymptomatic, indicating that the mutation had a high correlation with thrombosis. Patients in a South Korean study with p.Met252Arg mutation both had decreased AT activity, antigen level, and venous thrombosis.\n9\n We speculated that Met252 might be a hot spot mutation of the AT gene and carries a higher risk of thrombosis.\nIn order to further investigate the reasons for the reduction in AT levels caused by these three mutations, we ligated the coding sequence (CDS) of the AT gene into the eukaryotic expression vector to construct the AT‐WT plasmid. RT‐PCR results showed that the expression of mutant AT mRNA was not significantly different from that of the wild type, indicating that these three mutations had no significant effect on AT transcription. Western blot results showed that levels of AT protein of AT‐M252T and AT‐I421T in the cell lysates were significantly higher than that of AT‐WT, but the AT levels of both in the culture supernatant were significantly lower than that of AT‐WT, showing that the mutations of p.Met252Thr and p.Ile421Thr affected the normal secretion of AT protein from the inside to the outside of the cell, resulting in AT being retained in the cell, and causing the AT level to decrease. ELISA results showed the same findings, and further verified the retention of AT protein in the cell. Mutations such as AT‐Trp225Cys, AT‐Ala404Asp\n,23\n and AT‐Ala404Thr\n24\n had previously been reported. In vitro expression experiments showed that these mutations also led to impaired secretion of AT protein, type I inherited AT deficiency. For the p.Leu417Gln mutation, both ELISA and Western blot showed that the AT level expressed by the AT‐L417Q mutant was not significantly different from that of the wild type, which was consistent with the phenotypic test result of proband 2. The AT activity level of the patient was reduced, indicating that dysfunctional AT molecules may be generated by the mutation, thus impairing or damaging anticoagulant activity.\n25\n\n\nIn this article, we confirmed that pathogenic molecular mechanisms of p.Met252Thr, p.Ile421Thr, and p.Leu417Gln mutations led to inherited AT defects in three families. However, the genotype is not completely consistent with the clinical phenotype. Among the 25 people in the above‐mentioned three families, ten people carry AT mutations, but only five people displayed clinical manifestations; three probands had venous thrombosis, but only family 3 had a family history of hemorrhagic thrombosis. Family members in Family 1 and Family 2 who carried the same mutations as the proband had no thrombosis or related clinical manifestations, suggesting that patients with simple AT deficiency might not always develop thrombus. Therefore, it is necessary to carry out routine and specific antithrombotic prevention and treatment for such patients.","We reported three heterozygous missense mutations in the SERPINC1 gene, namely c.1358 T > C (p.Ile421Thr), c.1346 T > A (p.Leu417Gln), and c.851 T > C (p.Met252Thr). The pathogenic mechanisms of the three mutations were preliminarily explored through bioinformatics, protein modeling, and in vitro expression analysis, which enriched the AT genetic mutation database.","MSW and ML designed experiments, analyzed data, wrote the article, and supervised the project. SQL and STJ performed molecular modeling. SQL and YHJ designed experiments, conducted genetic analysis, and coagulation assays. All authors approved the final version of the article.","The authors declare no conflicts of interests.","Our study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China).","\nFigures S1–S4\n\nClick here for additional data file."],"string":"[\n \"Antithrombin (AT) is the major serine protease inhibitor (SERPIN) in plasma that regulates the proteolytic activities of coagulation proteases of both intrinsic and extrinsic pathways. AT is synthesized predominantly by hepatocytes, has a half‐life of approximately 2.4 days and a molecular weight of 58 kDa, and contains 432 amino acids.\\n1\\n AT is the major inhibitor of thrombin, activated coagulation factor IX (FIXa), and activated coagulation factor Х (FXa) in plasma, and is also able to inactivate the other serine proteases of the intrinsic coagulation pathway, activated factors XI and XII (FXIa and FXIIa).\\n2\\n In the physiologic state, AT circulates in a relatively inefficient inhibitor form. However, the anticoagulant effect of AT is greatly increased (at least 1000‐fold) by the presence of heparin and other heparin‐like glycosaminoglycans. This effect is the basis for the use of heparin and low‐molecular‐weight heparins (LMWH) as anticoagulants in patients with venous thromboembolism (VTE).\\n3\\n\\n\\nThe AT gene (SERPINC1) which encodes antithrombin is mapped on chromosome 1q23‐25 and comprises 7 exons encompassing 13.5 kb of genomic DNA.\\n4\\n The 1392 bp mRNA codes for a 464 amino acid polypeptide chain that undergoes several posttranslational modifications including cleavage of the signal peptide (32 amino acids) and N‐glycosylation.\\n5\\n, \\n6\\n Inherited AT deficiency is an autosomal‐dominant disorder caused by a defect in the SERPINC1 gene. It was first described in 1965 as one of the important risk factors for hemostatic diseases characterized by hereditary and recurrent thrombosis. In healthy people, the prevalence of inherited AT deficiency is around 0.02%–0.25%, but the incidence in patients with venous thrombosis is increased around 1%–2%.\\n7\\n, \\n8\\n, \\n9\\n Inherited AT deficiency is divided into type І with a reduction in the antigen levels and functional activity and type II deficiency with reduced antithrombin activity but almost normal antigen level.\\n10\\n, \\n11\\n\\n\\nMost cases with type I defects are heterozygous; homozygous patients have a higher risk of severe venous thromboembolism (VTE) in childhood and may even be fatal in utero.\\n9\\n For type II AT deficiency, the level of AT in the patient's plasma is not significantly affected; therefore, we speculate that type II has a slight thrombotic effect. Related studies have shown that patients with type II deficiency exhibit extensive and heterogeneous clinical phenotype.\\n12\\n\\n\\nInherited AT deficiency is widely considered to be one of the causes of thrombotic diseases. However, there are a few large cohort studies reported. More than 400 different mutations have been registered in the AT Mutation Database (The Human Gene Mutation Database, http://www.hgmd.cf.ac.uk/ac/all.php). Most are missense/nonsense mutations, the remaining being small insertion/deletion, large insertion/deletion, shear mutation, and complex mutation. The SERPINC1 gene mutations are heterogeneous in the population, and there are not many race‐specific mutations.\\n13\\n, \\n14\\n, \\n15\\n\\n\\nInherited AT deficiency is the main genetic risk factor for deep vein thrombosis (DVT) patients in China.\\n16\\n Sporadic cases with SERPINC1 gene mutations were reported in the population; however, no mutation hotspots were recognized. In this article, we presented three Chinese pedigrees with inherited AT deficiency, the phenotype, gene mutations, and expression of mutation sites in vitro. The molecular bases of the three pedigrees were analyzed.\",\n \"[SUBTITLE] Patients [SUBSECTION] Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\\nThree pedigrees of inherited AT deficiency\\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\\nProband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\\nThree pedigrees of inherited AT deficiency\\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\\n[SUBTITLE] Plasma AT assays [SUBSECTION] Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\\nPeripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\\n[SUBTITLE] \\nAT gene analysis [SUBSECTION] Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\\n17\\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\\nPeripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\\n17\\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\\n[SUBTITLE] Bioformatics analysis [SUBSECTION] The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\\nThe bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\\n[SUBTITLE] Construction, transfection, and purification [SUBSECTION] The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.\\nThe vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.\",\n \"Proband 1 was a 27‐year‐old woman who was admitted to the emergency department of our hospital for a sudden headache during puerperium, accompanied by nausea and vomiting. Magnetic resonance venography (MRV) showed thrombosis in the straight sinus, confluence of sinuses, and left transverse sinus. Seven family members of three generations had no relevant clinical manifestations. The pedigree is shown in Figure 1A.\\nThree pedigrees of inherited AT deficiency\\nProband 2 was a 74‐year‐old man, sent to our hospital for sudden left limb weakness and no movement of the upper and lower limbs. B‐ultrasound examination of the blood vessels of the lower extremities showed that there was thrombosis in the myenteric venous plexus of the legs. The pedigree is shown in Figure 1B.\\nProband 3 was a 30‐year‐old pregnant woman, admitted to our hospital because her left leg was swollen and painful. She had undergone a cesarean section 4 years ago and developed pulmonary embolism (PE) after that delivery. Her mother and uncle had the same history of thrombosis, and her grandfather died of PE at the age of 64, while the other family members had no relevant clinical manifestations. The pedigree is shown in Figure 1C.\\nIn order to establish a laboratory reference range, one hundred healthy subjects were randomly selected from our hospital as controls to exclude genetic polymorphisms; they included 50 males and 50 females with an average age of 28 years (from 16 to 52 years), without liver or kidney disease, no history of bleeding and thrombosis. Our study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China). All participants signed a written informed consent.\",\n \"Peripheral blood samples of three family members and 100 healthy subjects were collected in anticoagulant tubes and centrifuged at 2100 g for 10 min, to obtain upper platelet‐poor plasma (PPP) and lower blood cells. The upper plasma was examined by coagulation test, and the lower blood cells were used for genomic DNA extraction.\\nProthrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), and protein S activity (PS:A) were analyzed using the STAGO STA‐R‐Max automatic blood coagulometer (Diagnostica Stago). D‐dimers were measured by the immunoturbidimetric method, AT:A, and protein C activity (PC:A) by the chromogenic substrate method. All laboratory tests were performed in accordance with the manufacturer's instructions (STAGO Company). The AT:Ag were detected by enzyme‐linked immunosorbent assay (ELISA), using the Human Antithrombin‐III (MIM #613118) ELISA kit (Hangzhou Lianke Biotechnology Co., Ltd). The AT protein content was processed and analyzed by Western blot.\",\n \"Peripheral blood genomic DNA of all subjects was extracted from the anticoagulant blood samples using a DNA Extraction Kit (TIANGEN) according to the manufacturer's protocol. Primers for all 7 exons and their flanking sequences of SERPINC1 were designed with NCBI Primer. The specific amplification conditions were performed as previously described.\\n17\\n The amplified products were identified and sequenced (Shanghai Personal Biotechnology Co., Ltd.). The sequencing results were compared with the SERPINC1 standard sequence published in NCBI (GenBank ID: NG_012462.1), using Chromas software for mutation sites. Primers for R560Q and G20210A mutations which are common risk factors for hereditary venous thromboembolism were designed, and corresponding sites of the probands and their family members were screened.\",\n \"The bioinformatics tools, including Mutation Taster (http://www.mutationtaster.org/), PROVEAN (http:// provean.jcvi.org/index.php), LRT (http://varcards.biols.ac.cn/), SIFT (http://sift.jcvi.org), and PolyPhen‐2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), were applied to predict and analyze the possible impact of novel mutation on the structure and function of proteins.\\nConservation analysis relied on ClustalX‐2.1‐win (Science Foundation Ireland), a multiple sequence alignment software used to analyze the conservation of affected amino acids with nine homologous species (HomoloGene, http://www.ncbi.nlm.nih. gov/homologene).\\nThe PyMol software package (DeLano Scientific) was used to assess local structural effects of mutation in the AT gene, to determine the exact type and position of amino acid substitution in the three‐dimensional structural model. The AT protein crystal structure data were obtained from the Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/home.do, PDB ID: 1ANT).\",\n \"The vector used in this experiment was pCDH‐copGFP‐T2A‐Puro vector. Wild ‐type AT plasmid (AT‐WT) and empty vector were purchased from Nanjing Tsingke Biotechnology Co., Ltd. The AT‐WT was used as a template to construct AT‐M252T (p.Met252Thr), AT‐I421T (p.Ile421Thr), and AT‐L417Q (p.Leu417Gln) expression plasmids, according to manufacturer's instructions (QuikChange Lightning Site‐Directed Mutagenesis Kit). Mutant primers were designed using PrimerX online software (http://www.bioinformatics.org/primerx/) and synthesized by Nanjing Tsingke Biotechnology Co., Ltd.\\nHEK293FT cells were cultured in Dulbecco's modified Eagle medium (DMEM), which was supplemented with 10% fetal bovine serum (FBS). The wild‐type and mutant plasmids were transfected into human embryonic kidney 293 cells (HEK293FT). After 48 h transfection, total mRNA was extracted from the cells using TRIzol reagent and then reversely transcribed to cDNA. AT mRNA levels were determined using Magic SYBR Green qPCR Mix (Maibo, M223), according to manufacturer's instructions. The AT:Ag expression in cell culture supernatant and cell lysate was collected and analyzed by ELISA. Western blot was used to detect AT protein level in the cell supernatant and lysate.\",\n \"[SUBTITLE] Phenotype and genotype [SUBSECTION] In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\\nPhenotype results of three pedigrees with inherited AT deficiency\\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\\nIn pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\\nPhenotype results of three pedigrees with inherited AT deficiency\\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\\n[SUBTITLE] Bioformatics [SUBSECTION] The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\\n18\\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\\nBioinformatics prediction results of mutations\\n\\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\\nConservation analysis of homologous species of Met252, Leu417, and Ile421\\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\\nThe bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\\n18\\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\\nBioinformatics prediction results of mutations\\n\\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\\nConservation analysis of homologous species of Met252, Leu417, and Ile421\\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\\n[SUBTITLE] Transfection and expression [SUBSECTION] The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\\nThe AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\",\n \"In pedigree 1, the AT:A and AT:Ag of proband (II3), father (I1), and daughter (III1) all decreased; the values were 49%, 52 mg/dl, 48%, 42 mg/dl 40%, and 45 mg/dl, respectively, representing a type I AT deficiency. The phenotype results of the three pedigrees with inherited AT deficiency is shown in Table 1. Genetic analysis found that the above three members all carry the c.1358 T > C (Ile421Thr) heterozygous mutation in SERPINC1. In pedigree 2, the AT:A of proband (I1), and the eldest son (II2), second daughter (II4), and granddaughter (III3) decreased, with values 32%, 43%, 52%, and 48%, respectively; however, AT:Ag was within normal range, indicating a type II AT deficiency. They were all heterozygous for the c.1346 T > A (Leu417Gln) mutation. In pedigree 3, the AT:A and AT:Ag of proband (III2), mother (II3), and uncle (II1) showed a simultaneous decrease in AT:A and AT:Ag, with values of 38%, 44 mg/dl, 50%, 48 mg/dl, 40%, and 54 mg/dl, respectively, indicating a type I AT deficiency. Sequencing analysis revealed a heterozygous missense mutation c.851 T > C (p.Met252Thr) in the above three members of pedigree 3. The sequencing diagram is shown in Figure 2. Three corresponding sites in 100 healthy subjects were sequenced and excluded genetic polymorphisms. The FV‐Leiden and prothrombin G20210A of common thrombogenic gene mutations were screened and showed none existing in three probands and their family members.\\nPhenotype results of three pedigrees with inherited AT deficiency\\nSequencing results: (A) c. 1358 T wild type, (B) c. 1358 T > C heterozygous mutation of proband 1; (C) c. 1346 T wild type, (D) c.1346 T > A heterozygous mutation of proband 2; (E) c.851 T wild type, (F) c.851 T > C heterozygous mutation of proband 3\\nThe results of plasma AT protein from Western blot tests are shown in Figure 3. The plasma AT molecular weight of the three probands was consistent with the molecular weight of the normal mixed plasma control, but the plasma AT levels of probands 1 and 3 were significantly lower than those of the normal control. There was no difference between AT content of proband 2 and the normal control.\\nWB results of Plasma AT protein. The control was human mixed plasma. Results 1, 2, and 3 were proband 1, 2, and 3 respectively\",\n \"The bioinformatics prediction results are shown in Table 2. MutationTaster, PolyPhen‐2, PROVEAN, LRT, and SIFT all predicted the p.Met252Thr, and p.Ile421Thr mutations as “pathogenic and harmful”. For the p. Leu417Gln mutation, MutationTaster, PolyPhen‐2, and LRT suggested that it was harmful; SIFT and PROVEAN software labeled the mutations as “tolerable” and “neutral” respectively.\\n18\\n The ACMG guidelines classified p.Met252Thr and p.Ile421Thr mutations as “pathogenic,” and the p. Leu417Gln mutation as “likely pathogenic.” The conservation analysis results showed that the Met252, Leu417, and Ile421 were highly conserved among homologous species (Figure 4). In the p.Ile421Thr wild type, when Ile421 was replaced by Thr421, the original hydrogen bond was unchanged, but new hydrogen bonds were formed between Thr421 and Ile412, and Thr421 and Ile422, as shown in Figure 5. In the wild‐type AT protein, the main chain of Leu417 formed two hydrogen bonds with the side chain of Glu414 and Ser52. In the wild‐type AT model, there were two hydrogen bonds between the main chain of Met252 residues and Met320. When the Met252 was substituted by Thr, the original hydrogen bond remained, and an additional hydrogen bond was formed with Met320 (as shown in Figure 5).\\nBioinformatics prediction results of mutations\\n\\nNote: Meanings of scores are as follows: Mutation Taster: a probability close to 1 indicates a high predictive value; PolyPhen‐2: scores are evaluated as 0.000 (most probably benign) to 1.000 (most probably damaging); PROVEAN: the predefined threshold of score is 2.5, the score 2.5 or less (deleterious), greater than 2.5 (neutral); LRT: scores 0.01 or less (deleterious), greater than 0.01 (benign); SIFT: scores range from 0 to 1, 0.05 or less (damaging), greater than 0.05 (tolerated).\\nConservation analysis of homologous species of Met252, Leu417, and Ile421\\nMutation model analysis (A) p.Ile421Thr wild type, (B) p.Ile421Thr mutant type, (C) p.Met252Thr wild type, (D) p.Met252Thr mutant type. Hydrogen bonds are indicated by green dotted lines.\",\n \"The AT‐M252T, AT‐I421T, and AT‐L417Q mutant plasmids were successfully constructed by site‐directed mutation without other mutation sites verified by sequencing. The wild type and each mutant plasmid were observed under normal field and fluorescence field after 48 h transfection. The results are shown in Figures S1–S4. The wild‐type and each mutant plasmid were successfully transfected. The efficiency was approximately 85%.\\nThe mRNA expression level of each group was detected by the fluorescence quantitative RT‐PCR method; the results are shown in Figure S2. AT gene expression of the AT‐WT plasmid was 100%, AT‐M252T, AT‐I421T, and AT‐L417Q mutant were 94.65%, 94.62%, and 107.6%, respectively. There was no significant difference between these and AT‐WT in mRNA expression. The nontransfected control and samples transfected with empty vector did not express AT mRNA.\\nThe AT:Ag of the wild‐type and each mutant plasmid group were measured by ELISA. The cell culture supernatant and lysate of the wild‐type group were defined as 100%, and the calculated results were statistically analyzed by GraphPad 8.0 software as shown in Figure S3. The AT expression levels of AT‐M252T and AT‐I421T mutants in the cell culture supernatant were lower than those of wild‐type WT, while the expression levels of AT‐M252T and AT‐I421T mutants in the cell lysate were higher than that of wild‐type WT. The expression of AT of AT‐L417Q mutant in the cell culture supernatant and lysate was equivalent to that of the wild type, and there was no statistical difference.\\nSemi‐quantitative analyses of AT proteins were performed by Western blot in the cell lysate and culture supernatant. The results are shown in Figure S4, respectively. AT protein of AT‐M252T and AT‐I421T mutants expressed significantly more than the wild‐type in the cell lysate. AT protein expressions of AT‐M252T and AT‐I421T mutant were significantly reduced, compared with the wild type. AT protein expression of AT‐L417Q mutant levels were not much different from wild‐type cells in lysate and culture supernatant. The Western blot results of the three mutants were consistent with the ELISA results.\",\n \"Thrombophilia refers to a high thromboembolic tendency caused by certain risk factors, including hereditary and acquired factors. Genetic factors mainly include congenital defects of anticoagulant proteins such as AT, PC, and PS, mutations of prothrombin G20210A, and coagulation factor V Leiden.\\n19\\n Acquired risk factors include, but are not limited to, oral contraceptives, trauma, immobilization, and surgery, and other factors.\\n20\\n In this article, we presented 3 families with inherited AT deficiency and found 3 heterozygous missense variants, namely c.1358 T > C (p.Ile421Thr), c.1346 T > A (p.Leu417Gln), and c.851 T > C (p.Met252Thr). However, these three variants were not found in 100 healthy controls, thus excluding genetic polymorphisms.\\nProband 1 had p.Ile421Thr mutation and was diagnosed as a type I Inherited AT deficiency. The results of conservation analysis showed that the Ile421 site was highly conserved in homologous species, suggesting that this site plays an important role in the structure or function of AT protein. When le421 was replaced by Thr421, intermolecular hydrophobic force and hydrogen bonds changed, and the protein structure might have been affected, which was in accordance with the results of bioinformatics analysis. The p.Ile421Thr mutation has only been reported since 1994 by overseas researchers.\\n20\\n, \\n21\\n The overseas case presented recurrent venous thrombosis with a family history of thrombosis, and developed DVT and PE after the first pregnancy at the age of 27 years, with similar symptoms after the second and third pregnancies. Similar to our study, proband 1 was 27 years old and had venous thrombosis during puerperium, indicating that the p.Ile421Thr mutation has a higher risk of thrombosis, and when combined with acquired risk factors such as pregnancy and surgery, the p.Ile421Thr mutation has early onset characteristics. The p.Leu417Gln mutation was found in all four members in the second family (proband 2, his son, daughter, and granddaughter) who carried this heterozygous mutation; the details of which had been reported by our group.\\n18\\n All members in the third family including the proband 3, her mother, and uncle carried the p.Met252Thr heterozygous mutation. Western blot results showed that proband 3's plasma AT level was lower than the normal controls, in accordance with the phenotypic results, and diagnosed as a type I hereditary AT deficiency.\\n22\\n The bioinformatics software predicted that it would be pathogenic and that the protein structure may be affected. The hydrophobic interaction force and the number of hydrogen bonds between Thr252 and surrounding amino acids had been changed, which had an influence on the structural stability. In family 3, all three members carrying the heterozygous mutation of p.Met252Thr had a history of venous thrombosis; the wild‐type members were asymptomatic, indicating that the mutation had a high correlation with thrombosis. Patients in a South Korean study with p.Met252Arg mutation both had decreased AT activity, antigen level, and venous thrombosis.\\n9\\n We speculated that Met252 might be a hot spot mutation of the AT gene and carries a higher risk of thrombosis.\\nIn order to further investigate the reasons for the reduction in AT levels caused by these three mutations, we ligated the coding sequence (CDS) of the AT gene into the eukaryotic expression vector to construct the AT‐WT plasmid. RT‐PCR results showed that the expression of mutant AT mRNA was not significantly different from that of the wild type, indicating that these three mutations had no significant effect on AT transcription. Western blot results showed that levels of AT protein of AT‐M252T and AT‐I421T in the cell lysates were significantly higher than that of AT‐WT, but the AT levels of both in the culture supernatant were significantly lower than that of AT‐WT, showing that the mutations of p.Met252Thr and p.Ile421Thr affected the normal secretion of AT protein from the inside to the outside of the cell, resulting in AT being retained in the cell, and causing the AT level to decrease. ELISA results showed the same findings, and further verified the retention of AT protein in the cell. Mutations such as AT‐Trp225Cys, AT‐Ala404Asp\\n,23\\n and AT‐Ala404Thr\\n24\\n had previously been reported. In vitro expression experiments showed that these mutations also led to impaired secretion of AT protein, type I inherited AT deficiency. For the p.Leu417Gln mutation, both ELISA and Western blot showed that the AT level expressed by the AT‐L417Q mutant was not significantly different from that of the wild type, which was consistent with the phenotypic test result of proband 2. The AT activity level of the patient was reduced, indicating that dysfunctional AT molecules may be generated by the mutation, thus impairing or damaging anticoagulant activity.\\n25\\n\\n\\nIn this article, we confirmed that pathogenic molecular mechanisms of p.Met252Thr, p.Ile421Thr, and p.Leu417Gln mutations led to inherited AT defects in three families. However, the genotype is not completely consistent with the clinical phenotype. Among the 25 people in the above‐mentioned three families, ten people carry AT mutations, but only five people displayed clinical manifestations; three probands had venous thrombosis, but only family 3 had a family history of hemorrhagic thrombosis. Family members in Family 1 and Family 2 who carried the same mutations as the proband had no thrombosis or related clinical manifestations, suggesting that patients with simple AT deficiency might not always develop thrombus. Therefore, it is necessary to carry out routine and specific antithrombotic prevention and treatment for such patients.\",\n \"We reported three heterozygous missense mutations in the SERPINC1 gene, namely c.1358 T > C (p.Ile421Thr), c.1346 T > A (p.Leu417Gln), and c.851 T > C (p.Met252Thr). The pathogenic mechanisms of the three mutations were preliminarily explored through bioinformatics, protein modeling, and in vitro expression analysis, which enriched the AT genetic mutation database.\",\n \"MSW and ML designed experiments, analyzed data, wrote the article, and supervised the project. SQL and STJ performed molecular modeling. SQL and YHJ designed experiments, conducted genetic analysis, and coagulation assays. All authors approved the final version of the article.\",\n \"The authors declare no conflicts of interests.\",\n \"Our study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (China).\",\n \"\\nFigures S1–S4\\n\\nClick here for additional data file.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"materials-and-methods",null,null,null,null,null,"results",null,null,null,"discussion","conclusions",null,"COI-statement",null,"supplementary-material"],"string":"[\n null,\n \"materials-and-methods\",\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\",\n \"conclusions\",\n null,\n \"COI-statement\",\n null,\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["antithrombin","gene mutation","in vitro expression","inherited AT deficiency","thromboembolism"],"string":"[\n \"antithrombin\",\n \"gene mutation\",\n \"in vitro expression\",\n \"inherited AT deficiency\",\n \"thromboembolism\"\n]"}}},{"rowIdx":383,"cells":{"title":{"kind":"string","value":"Serum hsa_circ_0000615 is a prognostic biomarker of sorafenib resistance in hepatocellular carcinoma."},"pmid":{"kind":"string","value":"36268976"},"background_abstract":{"kind":"string","value":"Circular RNAs (circRNAs) can shape tumor progression and chemoresistance. How specific circRNAs shape hepatocellular carcinoma (HCC) chemoresistance, however, remains to be fully elucidated."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"In total, serum samples were collected from 202 HCC patients that had completed four sorafenib chemotherapy cycles. Serum hsa_circ_0000615 levels in these patients were quantified via quantitative real-time polymerase chain reaction (qRT-PCR), with demographic details and survival outcomes being recorded for subsequent analyses."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"We found hsa_circ_0000615 to be significantly upregulated in chemoresistant HCC patients relative to chemosensitive patients, with such upregulation being positively correlated with disease stage. Moreover, the area under the curve (AUC) value for hsa_circ_0000615 was moderately good, and high levels of hsa_circ_0000615 expression were associated with shorter overall survival among chemoresistant HCC patients."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Our results highlight hsa_circ_0000615 as a promising driver of sorafenib resistance in HCC patients, highlighting it as a promising target for the treatment of this deadly cancer type."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Carcinoma, Hepatocellular","RNA, Circular","Sorafenib","Liver Neoplasms","Prognosis","Biomarkers, Tumor"],"string":"[\n \"Humans\",\n \"Carcinoma, Hepatocellular\",\n \"RNA, Circular\",\n \"Sorafenib\",\n \"Liver Neoplasms\",\n \"Prognosis\",\n \"Biomarkers, Tumor\"\n]"},"pmcid":{"kind":"string","value":"9701853"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation [SUBSECTION] We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\nWe began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\n[SUBTITLE] Hsa_circ_0000615 levels are related to clinical features in HCC patients [SUBSECTION] Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\nNext, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\n[SUBTITLE] Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis [SUBSECTION] Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\nThrough Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\n[SUBTITLE] Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance [SUBSECTION] Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\nPrevious studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy."},"conclusions_title":{"kind":"string","value":"CONCLUSIONS"},"conclusions_text":{"kind":"string","value":"In summary, we herein found hsa_circ_0000615 upregulation to be prominent within serum samples from HCC patients, with such upregulation being significantly more pronounced in samples from chemosensitive patients relative to chemoresistant patients. As such, hsa_circ_0000615 is a promising target that warrants further study in an effort to understand the mechanistic basis for HCC patient chemoresistance."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Cell culture and clinical samples","Quantitative real‐time polymerase chain reaction","Statistical analysis","Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation","Hsa_circ_0000615 levels are related to clinical features in HCC patients","Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis","Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance"],"string":"[\n \"INTRODUCTION\",\n \"Cell culture and clinical samples\",\n \"Quantitative real‐time polymerase chain reaction\",\n \"Statistical analysis\",\n \"Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation\",\n \"Hsa_circ_0000615 levels are related to clinical features in HCC patients\",\n \"Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis\",\n \"Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance\"\n]"},"other_sections_texts":{"kind":"list like","value":["Hepatocellular carcinoma (HCC) is the leading cause of cancer‐related death worldwide.\n1\n In most cases, long‐term sorafenib administration is associated with the onset of chemoresistance. As such, there is a clear need to identify the mechanistic basis for sorafenib resistance and to design novel approaches to effectively treat this cancer type.\nCircular RNAs (circRNAs) are a covalently closed looping structure that renders them resistant to degradation and more stable than linear RNAs.\n2\n, \n3\n, \n4\n A growing body of evidence suggests that circRNAs can regulate a range of cancers and other important diseases by influencing cellular proliferation, survival, migration, glucose metabolism, and differentiation.\n5\n, \n6\n, \n7\n, \n8\n, \n9\n As such, circRNAs offer great promise as diagnostic biomarkers or therapeutic targets in cancer patients and individuals with other conditions.\n10\n\n\nRecently, circRNAs have been found to be dysregulated in HCC and to play an important functional role in this pathological context.\n11\n, \n12\n, \n13\n The recently identified circRNA hsa_circ_0000615 has been found to play an oncogenic role in prostate,\n14\n breast,\n15\n gastric,\n16\n and colorectal cancers.\n17\n Moreover, hsa_circ_0000615 has been found to promote HCC cell migration, invasion, stemness, and proliferation.\n18\n, \n19\n How hsa_circ_0000615 functions in the context of tumor chemoresistance, however, remains to be defined.\nHerein, we explored the expression of hsa_circ_0000615 in HCC patient serum and its relationship with patient clinical findings. Overall, we found that sorafenib‐resistant HCC patients exhibited hsa_circ_0000615 upregulation that was related to poorer overall survival (OS) outcomes. Moreover, hsa_circ_0000615 exhibited reasonably good area under the ROC curve (AUC) values, suggesting that it may offer value as a novel prognostic biomarker of sorafenib‐resistant HCC.","Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.","An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.","Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.","We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.","Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.","Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival","Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy."],"string":"[\n \"Hepatocellular carcinoma (HCC) is the leading cause of cancer‐related death worldwide.\\n1\\n In most cases, long‐term sorafenib administration is associated with the onset of chemoresistance. As such, there is a clear need to identify the mechanistic basis for sorafenib resistance and to design novel approaches to effectively treat this cancer type.\\nCircular RNAs (circRNAs) are a covalently closed looping structure that renders them resistant to degradation and more stable than linear RNAs.\\n2\\n, \\n3\\n, \\n4\\n A growing body of evidence suggests that circRNAs can regulate a range of cancers and other important diseases by influencing cellular proliferation, survival, migration, glucose metabolism, and differentiation.\\n5\\n, \\n6\\n, \\n7\\n, \\n8\\n, \\n9\\n As such, circRNAs offer great promise as diagnostic biomarkers or therapeutic targets in cancer patients and individuals with other conditions.\\n10\\n\\n\\nRecently, circRNAs have been found to be dysregulated in HCC and to play an important functional role in this pathological context.\\n11\\n, \\n12\\n, \\n13\\n The recently identified circRNA hsa_circ_0000615 has been found to play an oncogenic role in prostate,\\n14\\n breast,\\n15\\n gastric,\\n16\\n and colorectal cancers.\\n17\\n Moreover, hsa_circ_0000615 has been found to promote HCC cell migration, invasion, stemness, and proliferation.\\n18\\n, \\n19\\n How hsa_circ_0000615 functions in the context of tumor chemoresistance, however, remains to be defined.\\nHerein, we explored the expression of hsa_circ_0000615 in HCC patient serum and its relationship with patient clinical findings. Overall, we found that sorafenib‐resistant HCC patients exhibited hsa_circ_0000615 upregulation that was related to poorer overall survival (OS) outcomes. Moreover, hsa_circ_0000615 exhibited reasonably good area under the ROC curve (AUC) values, suggesting that it may offer value as a novel prognostic biomarker of sorafenib‐resistant HCC.\",\n \"Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\",\n \"An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\",\n \"Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.\",\n \"We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\",\n \"Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\",\n \"Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\",\n \"Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\\n20\\n gastric cancer,\\n21\\n and HCC.\\n22\\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","MATERIALS AND METHODS","Cell culture and clinical samples","Quantitative real‐time polymerase chain reaction","Statistical analysis","RESULTS","Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation","Hsa_circ_0000615 levels are related to clinical features in HCC patients","Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis","Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance","DISCUSSION","CONCLUSIONS","CONFLICT OF INTEREST"],"string":"[\n \"INTRODUCTION\",\n \"MATERIALS AND METHODS\",\n \"Cell culture and clinical samples\",\n \"Quantitative real‐time polymerase chain reaction\",\n \"Statistical analysis\",\n \"RESULTS\",\n \"Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation\",\n \"Hsa_circ_0000615 levels are related to clinical features in HCC patients\",\n \"Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis\",\n \"Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance\",\n \"DISCUSSION\",\n \"CONCLUSIONS\",\n \"CONFLICT OF INTEREST\"\n]"},"all_sections_texts":{"kind":"list like","value":["Hepatocellular carcinoma (HCC) is the leading cause of cancer‐related death worldwide.\n1\n In most cases, long‐term sorafenib administration is associated with the onset of chemoresistance. As such, there is a clear need to identify the mechanistic basis for sorafenib resistance and to design novel approaches to effectively treat this cancer type.\nCircular RNAs (circRNAs) are a covalently closed looping structure that renders them resistant to degradation and more stable than linear RNAs.\n2\n, \n3\n, \n4\n A growing body of evidence suggests that circRNAs can regulate a range of cancers and other important diseases by influencing cellular proliferation, survival, migration, glucose metabolism, and differentiation.\n5\n, \n6\n, \n7\n, \n8\n, \n9\n As such, circRNAs offer great promise as diagnostic biomarkers or therapeutic targets in cancer patients and individuals with other conditions.\n10\n\n\nRecently, circRNAs have been found to be dysregulated in HCC and to play an important functional role in this pathological context.\n11\n, \n12\n, \n13\n The recently identified circRNA hsa_circ_0000615 has been found to play an oncogenic role in prostate,\n14\n breast,\n15\n gastric,\n16\n and colorectal cancers.\n17\n Moreover, hsa_circ_0000615 has been found to promote HCC cell migration, invasion, stemness, and proliferation.\n18\n, \n19\n How hsa_circ_0000615 functions in the context of tumor chemoresistance, however, remains to be defined.\nHerein, we explored the expression of hsa_circ_0000615 in HCC patient serum and its relationship with patient clinical findings. Overall, we found that sorafenib‐resistant HCC patients exhibited hsa_circ_0000615 upregulation that was related to poorer overall survival (OS) outcomes. Moreover, hsa_circ_0000615 exhibited reasonably good area under the ROC curve (AUC) values, suggesting that it may offer value as a novel prognostic biomarker of sorafenib‐resistant HCC.","[SUBTITLE] Cell culture and clinical samples [SUBSECTION] Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\nHuman Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\n[SUBTITLE] Quantitative real‐time polymerase chain reaction [SUBSECTION] An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\nAn RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\n[SUBTITLE] Statistical analysis [SUBSECTION] Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.\nData are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.","Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.","An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.","Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.","[SUBTITLE] Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation [SUBSECTION] We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\nWe began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\n[SUBTITLE] Hsa_circ_0000615 levels are related to clinical features in HCC patients [SUBSECTION] Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\nNext, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\n[SUBTITLE] Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis [SUBSECTION] Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\nThrough Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\n[SUBTITLE] Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance [SUBSECTION] Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\nPrevious studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.","We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.","Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.","Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival","Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\n20\n gastric cancer,\n21\n and HCC.\n22\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.","Herein, we found that serum samples from HCC patients exhibited significant increases in hsa_circ_0000615 levels compared with those from control individuals. Moreover, the upregulation of this circRNA in samples from sorafenib‐resistant HCC patients relative to those from chemosensitive patients suggested that it may offer value as an independent predictor of patient outcomes.\nA growing body of evidence suggests that circRNAs are functionally important in cancer and may offer value as predictive biomarkers or therapeutic targets. In colorectal cancer, for example, circRNA_0000392 can promote tumor progression via the miR‐193a‐5p/PIK3R3/AKT axis.\n23\n Moreover, in non‐small cell lung cancer, circNDUFB2 can destabilize IGF2BPs and activate anti‐tumor immune responses to suppress tumor progression.\n24\n In HCC, circRNA‐SORE can stabilize YBX1 to drive sorafenib resistance.\n25\n As such, circRNAs function in a tumor‐specific manner.\nNeoadjuvant chemotherapy is a mainstay of treatment for many cancers and has been used with increasing frequency over the past decade, with sorafenib‐based neoadjuvant chemotherapy being a standard of care for HCC patients. Those HCC patients that undergo sorafenib‐based chemotherapy prior to radical cystectomy exhibit better OS outcomes, but a subset of patients fail to attain any benefit from such treatment, with pathological responses to neoadjuvant chemotherapy being predictive of disease‐specific survival outcomes. Identifying reliable biomarkers capable of guiding clinicians to the selection of patients most likely to benefit from chemotherapeutic intervention is thus a critical clinical task.\nIn HCC, circRNAs can function as central regulators of sorafenib‐resistance,\n26\n, \n27\n, \n28\n with hsa_circ_0000615 having previously been shown to drive HCC tumor growth and metastatic progression.\n18\n, \n19\n In this study, we further found hsa_circ_0000615 to be expressed at significantly higher levels in Hep G2/sorafenib and Huh 7/sorafenib cells relative to corresponding parental cell lines. The expression of this circRNA was similarly elevated in sorafenib‐resistant HCC patients compared with their chemosensitive counterparts, suggesting that hsa_circ_0000615 may offer value as a predictor of chemotherapeutic responses. Levels of hsa_circ_0000615 were also related to clinical stage, lymph node metastasis, and T stage in HCC patients, although they were unrelated to tumor histological stage, N stage, M stage, or patient age and gender. Kaplan–Meier analyses indicated that higher levels of hsa_circ_0000615 expression were associated with shorter patient OS compared with low levels of this circRNA. Moreover, chemoresistant HCC patients exhibited shorter OS and PFS compared with chemosensitive patients. Univariate and multivariate analyses further revealed clinical stage, T stage, lymph node metastasis, and chemoresistance to be correlated with OS and PFS outcomes, suggesting hsa_circ_0000615 to be a valuable independent predictor of HCC patient outcomes. In addition, the AUC value for this circRNA in HCC patients was 0.9238, indicating that serum levels of hsa_circ_0000615 can be used to reliably differentiate between HCC patients and healthy individuals.","In summary, we herein found hsa_circ_0000615 upregulation to be prominent within serum samples from HCC patients, with such upregulation being significantly more pronounced in samples from chemosensitive patients relative to chemoresistant patients. As such, hsa_circ_0000615 is a promising target that warrants further study in an effort to understand the mechanistic basis for HCC patient chemoresistance.","The authors of this work declare that they have no conflict of interest."],"string":"[\n \"Hepatocellular carcinoma (HCC) is the leading cause of cancer‐related death worldwide.\\n1\\n In most cases, long‐term sorafenib administration is associated with the onset of chemoresistance. As such, there is a clear need to identify the mechanistic basis for sorafenib resistance and to design novel approaches to effectively treat this cancer type.\\nCircular RNAs (circRNAs) are a covalently closed looping structure that renders them resistant to degradation and more stable than linear RNAs.\\n2\\n, \\n3\\n, \\n4\\n A growing body of evidence suggests that circRNAs can regulate a range of cancers and other important diseases by influencing cellular proliferation, survival, migration, glucose metabolism, and differentiation.\\n5\\n, \\n6\\n, \\n7\\n, \\n8\\n, \\n9\\n As such, circRNAs offer great promise as diagnostic biomarkers or therapeutic targets in cancer patients and individuals with other conditions.\\n10\\n\\n\\nRecently, circRNAs have been found to be dysregulated in HCC and to play an important functional role in this pathological context.\\n11\\n, \\n12\\n, \\n13\\n The recently identified circRNA hsa_circ_0000615 has been found to play an oncogenic role in prostate,\\n14\\n breast,\\n15\\n gastric,\\n16\\n and colorectal cancers.\\n17\\n Moreover, hsa_circ_0000615 has been found to promote HCC cell migration, invasion, stemness, and proliferation.\\n18\\n, \\n19\\n How hsa_circ_0000615 functions in the context of tumor chemoresistance, however, remains to be defined.\\nHerein, we explored the expression of hsa_circ_0000615 in HCC patient serum and its relationship with patient clinical findings. Overall, we found that sorafenib‐resistant HCC patients exhibited hsa_circ_0000615 upregulation that was related to poorer overall survival (OS) outcomes. Moreover, hsa_circ_0000615 exhibited reasonably good area under the ROC curve (AUC) values, suggesting that it may offer value as a novel prognostic biomarker of sorafenib‐resistant HCC.\",\n \"[SUBTITLE] Cell culture and clinical samples [SUBSECTION] Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\\nHuman Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\\n[SUBTITLE] Quantitative real‐time polymerase chain reaction [SUBSECTION] An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\\nAn RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\\n[SUBTITLE] Statistical analysis [SUBSECTION] Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.\\nData are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.\",\n \"Human Hep G2 and Huh 7 were grown in DMEM (Invitrogen, NY, USA). Hep G2/sorafenib and Huh 7/sorafenib cell lines were established by maintaining Hep G2 and Huh 7 cells at 1 mmol/L sorafenib and gradually increasing it at a rate of 0.5 mmol/L per month (up to 5 mmol/L) more than 10‐month. Serum samples from 202 HCC patients and 202 healthy controls were obtained from the First Affiliated Hospital of Bengbu Medical College. This study was approved by the Ethics Committee of the First Affiliated Hospital of Bengbu Medical College, with all patients having provided written informed consent.\",\n \"An RNA Isolation Kit (Vazyme Biotech, Nanjing, China) was used to extract total RNA from 500 μl of patient serum, after which a Prime Script RT reagent Kit (Takara, Dalian, China) was used for cDNA synthesis. Prepared cDNA was then used as input for qPCR reactions performed with SYBR Green (Takara). The U6 small nuclear B noncoding RNA (U6) was used to normalize expression values via the 2−ΔΔCt method, with primers used being as follows: hsa_circ_0000615: F 5′–CAGCGCTATCCTTTGGGA–3′, R 5′–GACCTGCCACATTGGTCAGTA–3′; U6: F 5′–TGCGGGTGCTCGCTTCGGCAGC–3′, R 5′–GTGCAGGGTCCGAGGT–3′.\",\n \"Data are means ± standard deviation (SD) and were compared via Student's t tests using GraphPad Prism 7. The Kaplan–Meier method was used for survival analyses, with p < .05 as the threshold of significance.\",\n \"[SUBTITLE] Hepatocellular carcinoma patients exhibit serum hsa_circ_0000615 upregulation [SUBSECTION] We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\\nWe began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\\n[SUBTITLE] Hsa_circ_0000615 levels are related to clinical features in HCC patients [SUBSECTION] Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\\nNext, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\\n[SUBTITLE] Hsa_circ_0000615 is associated with poor chemoresistant hepatocellular carcinoma patient prognosis [SUBSECTION] Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\\nThrough Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\\n[SUBTITLE] Serum hsa_circ_0000615 levels offer diagnostic utility for the detection of hepatocellular carcinoma chemoresistance [SUBSECTION] Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\\n20\\n gastric cancer,\\n21\\n and HCC.\\n22\\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\\nPrevious studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\\n20\\n gastric cancer,\\n21\\n and HCC.\\n22\\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\",\n \"We began by assessing the levels of hsa_circ_0000615 in control and sorafenib‐resistant HCC cells. The sorafenib‐resistant Hep G2/sorafenib and Huh 7/sorafenib cell lines exhibited marked upregulation of this circRNA relative to corresponding parental cell lines (Figure 1A). To explore the potential utility of hsa_circ_0000615 as a biomarker of chemoresistance, we then assessed the levels of this circRNA in serum samples from 202 HCC patients and 202 healthy controls. HCC patients exhibited significantly elevated serum hsa_circ_0000615 levels compared with healthy controls (Figure 1B). Moreover, hsa_circ_0000615 expression levels were higher in sorafenib‐resistant patients (n = 122) relative to those in sorafenib‐sensitive individuals (n = 80) (Figure 1C). As such, hsa_circ_0000615 offers potential value as an HCC chemotherapy biomarker.\\nHepatocellular carcinoma (HCC) patient serum samples exhibit hsa_circ_0000615 upregulation. (A). Hsa_circ_0000615 levels were significantly increased in sorafenib‐resistant HCC cells. (B). Serum hsa_circ_0000615 levels were higher in HCC patients. (C). sorafenib‐resistant patients (n = 122) exhibited higher levels of hsa_circ_0000615 expression compared with sorafenib‐sensitive patients in the before treatment and after treatment (n = 80). *p < .05.\",\n \"Next, we stratified HCC patients into those with high and low levels of serum hsa_circ_0000615 based on the mean expression of this circRNA in this cohort and compared clinical features between these two patient groups. Chi‐squared analyses revealed that hsa_circ_0000615 expression was associated with clinical stage, TNM stage, and lymph node metastasis (Table 1), but was unrelated to age, sex, or histological grade. Kaplan–Meier analyses indicated that patients exhibiting higher levels of hsa_circ_0000615 expression presented with shorter OS relative to patients expressing low hsa_circ_0000615 levels (Figure 2).\\nCorrelations between hsa_circ_0000615 levels and hepatocellular carcinoma patient clinicopathological features\\nHsa_circ_0000615 levels are linked with hepatocellular carcinoma patient survival. Patients exhibiting higher hsa_circ_0000615 expression levels exhibited prolonged overall survival compared with patients with lower levels of this circRNA.\",\n \"Through Kaplan–Meier analyses and log‐rank tests, we found chemoresistant HCC patients to exhibit significantly reduced OS and progression‐free survival (PFS) compared with chemosensitive patients (Figure 3). Through Cox proportional hazards regression analyses, we determined that clinical stage, chemoresistance, TNM stage, lymph node metastasis, and hsa_circ_0000615 levels were associated with patient PFS (Table 2) and OS (Table 3), highlighting hsa_circ_0000615 as a promising independent predictor of chemoresistant HCC patient survival.\\nHsa_circ_0000615 levels were significantly linked to poor outcomes in chemoresistant hepatocellular carcinoma (HCC) patients. Chemoresistant HCC patients exhibited significantly decreased progression‐free survival (A) and overall survival (B) relative to chemosensitive patients.\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient progression‐free survival\\nUnivariate and multivariate analyses of hepatocellular carcinoma patient overall survival\",\n \"Previous studies have shown that circRNAs show excellent potential diagnostic utility in various cancers, such as, breast cancer,\\n20\\n gastric cancer,\\n21\\n and HCC.\\n22\\n To assess the potential diagnostic utility of serum hsa_circ_0000615 in patients with HCC, the area under the receiver operating characteristic (ROC) curve (AUC) was determined and found to be 0.9238 (95% CI, 0.8915–0.956, Figure 4, p < .0001), consistent with the value of serum hsa_circ_0000615 as a biomarker capable of differentiating between HCC patients and healthy controls.\\nSerum hsa_circ_0000615 levels offer diagnostic value as predictors of hepatocellular carcinoma (HCC) patient chemoresistance. Receiver‐operating characteristic curves were used to differentiate between chemoresistant HCC patients before and after therapy.\",\n \"Herein, we found that serum samples from HCC patients exhibited significant increases in hsa_circ_0000615 levels compared with those from control individuals. Moreover, the upregulation of this circRNA in samples from sorafenib‐resistant HCC patients relative to those from chemosensitive patients suggested that it may offer value as an independent predictor of patient outcomes.\\nA growing body of evidence suggests that circRNAs are functionally important in cancer and may offer value as predictive biomarkers or therapeutic targets. In colorectal cancer, for example, circRNA_0000392 can promote tumor progression via the miR‐193a‐5p/PIK3R3/AKT axis.\\n23\\n Moreover, in non‐small cell lung cancer, circNDUFB2 can destabilize IGF2BPs and activate anti‐tumor immune responses to suppress tumor progression.\\n24\\n In HCC, circRNA‐SORE can stabilize YBX1 to drive sorafenib resistance.\\n25\\n As such, circRNAs function in a tumor‐specific manner.\\nNeoadjuvant chemotherapy is a mainstay of treatment for many cancers and has been used with increasing frequency over the past decade, with sorafenib‐based neoadjuvant chemotherapy being a standard of care for HCC patients. Those HCC patients that undergo sorafenib‐based chemotherapy prior to radical cystectomy exhibit better OS outcomes, but a subset of patients fail to attain any benefit from such treatment, with pathological responses to neoadjuvant chemotherapy being predictive of disease‐specific survival outcomes. Identifying reliable biomarkers capable of guiding clinicians to the selection of patients most likely to benefit from chemotherapeutic intervention is thus a critical clinical task.\\nIn HCC, circRNAs can function as central regulators of sorafenib‐resistance,\\n26\\n, \\n27\\n, \\n28\\n with hsa_circ_0000615 having previously been shown to drive HCC tumor growth and metastatic progression.\\n18\\n, \\n19\\n In this study, we further found hsa_circ_0000615 to be expressed at significantly higher levels in Hep G2/sorafenib and Huh 7/sorafenib cells relative to corresponding parental cell lines. The expression of this circRNA was similarly elevated in sorafenib‐resistant HCC patients compared with their chemosensitive counterparts, suggesting that hsa_circ_0000615 may offer value as a predictor of chemotherapeutic responses. Levels of hsa_circ_0000615 were also related to clinical stage, lymph node metastasis, and T stage in HCC patients, although they were unrelated to tumor histological stage, N stage, M stage, or patient age and gender. Kaplan–Meier analyses indicated that higher levels of hsa_circ_0000615 expression were associated with shorter patient OS compared with low levels of this circRNA. Moreover, chemoresistant HCC patients exhibited shorter OS and PFS compared with chemosensitive patients. Univariate and multivariate analyses further revealed clinical stage, T stage, lymph node metastasis, and chemoresistance to be correlated with OS and PFS outcomes, suggesting hsa_circ_0000615 to be a valuable independent predictor of HCC patient outcomes. In addition, the AUC value for this circRNA in HCC patients was 0.9238, indicating that serum levels of hsa_circ_0000615 can be used to reliably differentiate between HCC patients and healthy individuals.\",\n \"In summary, we herein found hsa_circ_0000615 upregulation to be prominent within serum samples from HCC patients, with such upregulation being significantly more pronounced in samples from chemosensitive patients relative to chemoresistant patients. As such, hsa_circ_0000615 is a promising target that warrants further study in an effort to understand the mechanistic basis for HCC patient chemoresistance.\",\n \"The authors of this work declare that they have no conflict of interest.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"materials-and-methods",null,null,null,"results",null,null,null,null,"discussion","conclusions","COI-statement"],"string":"[\n null,\n \"materials-and-methods\",\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n \"discussion\",\n \"conclusions\",\n \"COI-statement\"\n]"},"keywords":{"kind":"list like","value":["hepatocellular carcinoma","hsa_circ_0000615","sorafenib resistance"],"string":"[\n \"hepatocellular carcinoma\",\n \"hsa_circ_0000615\",\n \"sorafenib resistance\"\n]"}}},{"rowIdx":384,"cells":{"title":{"kind":"string","value":"Low plasma growth/differentiation factor 1 levels are associated with liver fibrosis in patients with stable angina."},"pmid":{"kind":"string","value":"36268984"},"background_abstract":{"kind":"string","value":"Growth differentiation factor 1 (GDF1) is a member of the transforming growth factor-β (TGF-β) superfamily and a protective mediator against the development of post-infarction cardiac remodeling by negatively regulating MEK-ERK1/2 and Smad signaling pathways in the heart. The TGF-β/SMAD pathway has been shown to play a key role in the development of hepatic fibrosis. In addition, fatty liver disease has been associated with reduced MEK/ERK1/2 signaling. However, no previous study has investigated the association between GDF1 and liver fibrosis. Therefore, the aim of this study was to investigate the association between plasma GDF1 and liver fibrosis in patients with stable angina."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We included 327 consecutive patients with stable angina. ELISA was used to measure circulating levels of GDF1, and the fibrosis-4 index was used to assess liver fibrosis."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The advanced liver fibrosis group had lower median plasma GDF1 levels than those with minimal liver fibrosis. There was a significant negative association between GDF1 plasma level and fibrosis-4 index (r = -0.135, p = 0.019). A lower concentration of GDF1 was significantly and independently associated with an increased risk of liver fibrosis when concentration was analyzed as a continuous variable and by tertile. In addition, fibrosis-4 index, aspartate aminotransferase (AST)-to-platelet ratio index, and AST/alanine aminotransferase ratio were significantly associated with GDF1 concentration."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Our results indicated an association between low plasma GDF1 and liver fibrosis in the enrolled patients. Further investigations into the role of plasma GDF1 in the pathogenesis of liver fibrosis are warranted."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Angina, Stable","Growth Differentiation Factor 1","Liver","Liver Cirrhosis","Mitogen-Activated Protein Kinase Kinases","Smad Proteins","Transforming Growth Factor beta"],"string":"[\n \"Humans\",\n \"Angina, Stable\",\n \"Growth Differentiation Factor 1\",\n \"Liver\",\n \"Liver Cirrhosis\",\n \"Mitogen-Activated Protein Kinase Kinases\",\n \"Smad Proteins\",\n \"Transforming Growth Factor beta\"\n]"},"pmcid":{"kind":"string","value":"9701856"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Characteristics of the patients by liver fibrosis category [SUBSECTION] Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.\nTable 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.\n[SUBTITLE] Association between plasma GDF1 and liver fibrosis [SUBSECTION] We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\nWe found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\n[SUBTITLE] Associations between plasma GDF1 level and clinical laboratory data [SUBSECTION] Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\nSimple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase."},"conclusions_title":{"kind":"string","value":"CONCLUSIONS"},"conclusions_text":{"kind":"string","value":"Our results indicated that in patients with stable angina, low plasma GDF1 was associated with liver fibrosis. Further studies are warranted to investigate the association between plasma GDF1 and the pathogenesis of liver fibrosis."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Patient population","Baseline data collection","Laboratory measurements","Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP)","Definitions","Statistical analysis","Characteristics of the patients by liver fibrosis category","Association between plasma GDF1 and liver fibrosis","Associations between plasma GDF1 level and clinical laboratory data","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","INFORMED CONSENT"],"string":"[\n \"INTRODUCTION\",\n \"Patient population\",\n \"Baseline data collection\",\n \"Laboratory measurements\",\n \"Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP)\",\n \"Definitions\",\n \"Statistical analysis\",\n \"Characteristics of the patients by liver fibrosis category\",\n \"Association between plasma GDF1 and liver fibrosis\",\n \"Associations between plasma GDF1 level and clinical laboratory data\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"INFORMED CONSENT\"\n]"},"other_sections_texts":{"kind":"list like","value":["Cardiovascular disease (CVD) is the leading cause of death worldwide,\n1\n and is precipitated by cardiometabolic risk factors including chronic liver disease, diabetes, metabolic syndrome, and chronic renal disease.\n1\n, \n2\n, \n3\n Chronic liver disease is a progressive deterioration of liver function, and the most common causes are hepatitis, alcohol abuse, and nonalcoholic fatty liver disease (NAFLD).\n4\n NAFLD is common in patients with CVD, and both diseases share common risks factor such as high cholesterol intake, alcohol consumption, diabetes and other metabolic abnormalities.\n5\n In addition, recent studies have reported associations between liver fibrosis, which is the irreversible progression of NAFLD, and coronary artery disease (CAD), cardiac arrhythmias, and impaired heart function.\n3\n, \n6\n, \n7\n Furthermore, previous study also revealed that higher liver fibrosis scores are associated with increased risks of cardiovascular and all‐cause mortality among patients with CAD. Liver fibrosis scores might play a potential role in CAD prognosis prediction.\n3\n However, the underlying mechanisms and interactions remain unknown.\nGrowth differentiation factors (GDFs) are proteins which belong to the transforming growth factor‐β (TGF‐β) superfamily, and 15 GDFs have been identified to date (GDF1 to GDF15).\n8\n The TGF‐β superfamily comprised structurally related polypeptides which have been shown to regulate cell differentiation and growth in various adult and embryonic tissues. GDF1 has been shown to be specifically expressed in the nervous systems of adult mice and late‐stage embryos.\n9\n, \n10\n A previous study showed that GDF1 had favorable effects on vascular endothelial function, sodium excretion, and post‐infarction cardiac remodeling in an animal study and also in patients with type 2 diabetes and normal individuals.\n11\n The underlying mechanism is believed to be through the negative regulation of MEK‐ERK1/2 and Smad signaling pathways.\n11\n Interestingly, the TGF‐β/SMAD pathway has also been found to play a key role in the development of hepatic fibrosis in subjects with fatty liver disease via suppression of the MEK/ERK1/2 signaling pathway.\n12\n, \n13\n However, few studies have investigated the relationship between plasma GDF1 level and liver fibrosis. Therefore, the aim of this study was to evaluate the relationship between plasma GDF1 level and liver fibrosis in a cohort of Taiwanese patients with stable angina.","We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\n14\n, \n15\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\n16\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.","A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.","Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\n17\n, \n18\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\n19\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\n20\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).","After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.","We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\n21\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\n22\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\n20\n, \n23\n, \n24\n\n","Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).","Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.","We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.","Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.","All authors contributed to this study. W.‐C.H., W.‐H.T., and C.‐C.H. conceived and designed the study. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. provided the methodology. F.‐M.C. performed the formal analysis, and project administration. T.‐H.Y., C.‐C.W., and C.‐C.H. validated the data. T.‐H.Y., C.‐C.W., W.‐C.H., and C.‐P.W. performed the investigation, resources, and data curation. T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. prepared the manuscript. W.‐H.T., T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. reviewed and edited the manuscript. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. performed the visualization. W.‐C.H. and C.‐C.H. performed the supervision and funding acquisition. All authors have read and agreed to the published version of the manuscript.","E‐Da Hospital financially supported this research under Contracts EDAHP109002, EDAHI109002, and EDAHI110001.","Each patient provided written informed consent before being enrolled into the study."],"string":"[\n \"Cardiovascular disease (CVD) is the leading cause of death worldwide,\\n1\\n and is precipitated by cardiometabolic risk factors including chronic liver disease, diabetes, metabolic syndrome, and chronic renal disease.\\n1\\n, \\n2\\n, \\n3\\n Chronic liver disease is a progressive deterioration of liver function, and the most common causes are hepatitis, alcohol abuse, and nonalcoholic fatty liver disease (NAFLD).\\n4\\n NAFLD is common in patients with CVD, and both diseases share common risks factor such as high cholesterol intake, alcohol consumption, diabetes and other metabolic abnormalities.\\n5\\n In addition, recent studies have reported associations between liver fibrosis, which is the irreversible progression of NAFLD, and coronary artery disease (CAD), cardiac arrhythmias, and impaired heart function.\\n3\\n, \\n6\\n, \\n7\\n Furthermore, previous study also revealed that higher liver fibrosis scores are associated with increased risks of cardiovascular and all‐cause mortality among patients with CAD. Liver fibrosis scores might play a potential role in CAD prognosis prediction.\\n3\\n However, the underlying mechanisms and interactions remain unknown.\\nGrowth differentiation factors (GDFs) are proteins which belong to the transforming growth factor‐β (TGF‐β) superfamily, and 15 GDFs have been identified to date (GDF1 to GDF15).\\n8\\n The TGF‐β superfamily comprised structurally related polypeptides which have been shown to regulate cell differentiation and growth in various adult and embryonic tissues. GDF1 has been shown to be specifically expressed in the nervous systems of adult mice and late‐stage embryos.\\n9\\n, \\n10\\n A previous study showed that GDF1 had favorable effects on vascular endothelial function, sodium excretion, and post‐infarction cardiac remodeling in an animal study and also in patients with type 2 diabetes and normal individuals.\\n11\\n The underlying mechanism is believed to be through the negative regulation of MEK‐ERK1/2 and Smad signaling pathways.\\n11\\n Interestingly, the TGF‐β/SMAD pathway has also been found to play a key role in the development of hepatic fibrosis in subjects with fatty liver disease via suppression of the MEK/ERK1/2 signaling pathway.\\n12\\n, \\n13\\n However, few studies have investigated the relationship between plasma GDF1 level and liver fibrosis. Therefore, the aim of this study was to evaluate the relationship between plasma GDF1 level and liver fibrosis in a cohort of Taiwanese patients with stable angina.\",\n \"We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\\n14\\n, \\n15\\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\\n16\\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\",\n \"A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\",\n \"Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\\n17\\n, \\n18\\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\\n19\\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\\n20\\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\",\n \"After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\",\n \"We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\\n21\\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\\n22\\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\\n20\\n, \\n23\\n, \\n24\\n\\n\",\n \"Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).\",\n \"Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\\n\\nNote: Data are mean ± SD or frequency (percentage).\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\\nSignificant difference was tested using log‐transformed data.\",\n \"We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\\n\\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\\n\\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\\nAbbreviation: GDF1, growth/differentiation factor 1.\\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\",\n \"Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\\nRegression coefficient adjusted for age and sex.\\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\",\n \"All authors contributed to this study. W.‐C.H., W.‐H.T., and C.‐C.H. conceived and designed the study. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. provided the methodology. F.‐M.C. performed the formal analysis, and project administration. T.‐H.Y., C.‐C.W., and C.‐C.H. validated the data. T.‐H.Y., C.‐C.W., W.‐C.H., and C.‐P.W. performed the investigation, resources, and data curation. T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. prepared the manuscript. W.‐H.T., T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. reviewed and edited the manuscript. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. performed the visualization. W.‐C.H. and C.‐C.H. performed the supervision and funding acquisition. All authors have read and agreed to the published version of the manuscript.\",\n \"E‐Da Hospital financially supported this research under Contracts EDAHP109002, EDAHI109002, and EDAHI110001.\",\n \"Each patient provided written informed consent before being enrolled into the study.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","MATERIALS AND METHODS","Patient population","Baseline data collection","Laboratory measurements","Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP)","Definitions","Statistical analysis","RESULTS","Characteristics of the patients by liver fibrosis category","Association between plasma GDF1 and liver fibrosis","Associations between plasma GDF1 level and clinical laboratory data","DISCUSSION","CONCLUSIONS","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","CONFLICT OF INTEREST","INFORMED CONSENT"],"string":"[\n \"INTRODUCTION\",\n \"MATERIALS AND METHODS\",\n \"Patient population\",\n \"Baseline data collection\",\n \"Laboratory measurements\",\n \"Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP)\",\n \"Definitions\",\n \"Statistical analysis\",\n \"RESULTS\",\n \"Characteristics of the patients by liver fibrosis category\",\n \"Association between plasma GDF1 and liver fibrosis\",\n \"Associations between plasma GDF1 level and clinical laboratory data\",\n \"DISCUSSION\",\n \"CONCLUSIONS\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"CONFLICT OF INTEREST\",\n \"INFORMED CONSENT\"\n]"},"all_sections_texts":{"kind":"list like","value":["Cardiovascular disease (CVD) is the leading cause of death worldwide,\n1\n and is precipitated by cardiometabolic risk factors including chronic liver disease, diabetes, metabolic syndrome, and chronic renal disease.\n1\n, \n2\n, \n3\n Chronic liver disease is a progressive deterioration of liver function, and the most common causes are hepatitis, alcohol abuse, and nonalcoholic fatty liver disease (NAFLD).\n4\n NAFLD is common in patients with CVD, and both diseases share common risks factor such as high cholesterol intake, alcohol consumption, diabetes and other metabolic abnormalities.\n5\n In addition, recent studies have reported associations between liver fibrosis, which is the irreversible progression of NAFLD, and coronary artery disease (CAD), cardiac arrhythmias, and impaired heart function.\n3\n, \n6\n, \n7\n Furthermore, previous study also revealed that higher liver fibrosis scores are associated with increased risks of cardiovascular and all‐cause mortality among patients with CAD. Liver fibrosis scores might play a potential role in CAD prognosis prediction.\n3\n However, the underlying mechanisms and interactions remain unknown.\nGrowth differentiation factors (GDFs) are proteins which belong to the transforming growth factor‐β (TGF‐β) superfamily, and 15 GDFs have been identified to date (GDF1 to GDF15).\n8\n The TGF‐β superfamily comprised structurally related polypeptides which have been shown to regulate cell differentiation and growth in various adult and embryonic tissues. GDF1 has been shown to be specifically expressed in the nervous systems of adult mice and late‐stage embryos.\n9\n, \n10\n A previous study showed that GDF1 had favorable effects on vascular endothelial function, sodium excretion, and post‐infarction cardiac remodeling in an animal study and also in patients with type 2 diabetes and normal individuals.\n11\n The underlying mechanism is believed to be through the negative regulation of MEK‐ERK1/2 and Smad signaling pathways.\n11\n Interestingly, the TGF‐β/SMAD pathway has also been found to play a key role in the development of hepatic fibrosis in subjects with fatty liver disease via suppression of the MEK/ERK1/2 signaling pathway.\n12\n, \n13\n However, few studies have investigated the relationship between plasma GDF1 level and liver fibrosis. Therefore, the aim of this study was to evaluate the relationship between plasma GDF1 level and liver fibrosis in a cohort of Taiwanese patients with stable angina.","[SUBTITLE] Patient population [SUBSECTION] We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\n14\n, \n15\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\n16\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\nWe enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\n14\n, \n15\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\n16\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\n[SUBTITLE] Baseline data collection [SUBSECTION] A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\nA detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\n[SUBTITLE] Laboratory measurements [SUBSECTION] Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\n17\n, \n18\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\n19\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\n20\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\nFasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\n17\n, \n18\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\n19\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\n20\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\n[SUBTITLE] Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP) [SUBSECTION] After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\nAfter obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\n[SUBTITLE] Definitions [SUBSECTION] We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\n21\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\n22\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\n20\n, \n23\n, \n24\n\n\nWe used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\n21\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\n22\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\n20\n, \n23\n, \n24\n\n\n[SUBTITLE] Statistical analysis [SUBSECTION] Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).\nContinuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).","We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\n14\n, \n15\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\n16\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.","A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.","Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\n17\n, \n18\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\n19\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\n20\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).","After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.","We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\n21\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\n22\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\n20\n, \n23\n, \n24\n\n","Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).","[SUBTITLE] Characteristics of the patients by liver fibrosis category [SUBSECTION] Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.\nTable 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.\n[SUBTITLE] Association between plasma GDF1 and liver fibrosis [SUBSECTION] We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\nWe found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\n[SUBTITLE] Associations between plasma GDF1 level and clinical laboratory data [SUBSECTION] Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\nSimple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.","Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\n\nNote: Data are mean ± SD or frequency (percentage).\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\nSignificant difference was tested using log‐transformed data.","We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\n\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\n\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\nAbbreviation: GDF1, growth/differentiation factor 1.\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.","Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\nRegression coefficient adjusted for age and sex.\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.","According to a previous clinical study, fatty liver disease is common in patients with CAD, and it can progress to liver fibrosis resulting in higher rates of all‐cause and cardiovascular mortality.\n3\n Similarly, in the current study, we found that among the 327 stable angina patients, more than 75% had moderate to advanced liver fibrosis (Table 1). In the past decade, several studies have evaluated the relationship between fatty liver disease and CAD. Although some chronic liver diseases such as NAFLD share similar risk factors with CAD (such as hyperglycemia, insulin resistance, hypertension, inflammation, dyslipidemia, hyperuricemia, hypoadiponectinemia, renal failure, and obesity),\n25\n, \n26\n, \n27\n the underlying molecular relationship between chronic liver disease and CAD is unknown. Clinically, a meta‐analysis found that NAFLD was associated with the severity of atherosclerosis, including coronary calcification levels, carotid intima‐media thickness, arterial stiffness and endothelial dysfunction.\n28\n Furthermore, the review by Meex et al. in 2017 also showed that many hepatokines such as fetuin A, fetuin B, retinol‐binding protein 4, and selenoprotein P are involved in both fatty liver disease and insulin resistance, which in turn influences the process of atherosclerosis.\n29\n However, no previous report has described the underlying molecular pathogenesis between chronic liver disease and CAD. To the best of our knowledge, this is the first study to show that GDF1, which is involved in both post‐infarction cardiac remodeling and the development of fatty liver disease, was associated with liver fibrosis in patients with stable angina.\nThe cytokine system is very complex. Maintaining a balanced healthy environment for all human cells is difficult, and a factor which is beneficial for certain circumstances may be harmful for another. GDF1 is a member of the TGF‐β superfamily and is known to be a protective mediator against the development of post‐infarction cardiac remodeling via negative regulation of the Smad signaling and MEK‐ERK1/2 pathways in the heart.\n11\n, \n30\n, \n31\n A suppressed MEK/ERK1/2 signaling pathway has been associated with fatty liver disease, liver fibrosis, and poor hepatoma differentiation in recent studies.\n12\n, \n13\n In this study, a lower plasma GDF1 concentration was significantly independently associated with and showed a significant linear trend with liver fibrosis (Tables 2 and 3). It is reasonable that a decrease in GDF1 could enhance the progress of liver fibrosis. In addition, the Gensini score and number of diseased coronary arteries, which was used to evaluate CAD severity, did not show a significant difference among tertiles of GDF1 (data not shown). The role of GDF1 post infarction is known, however it remains unclear in CAD.\nIn analysis of the associations between common risk factors for liver fibrosis, we found no significant differences in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, LDL‐C, HDL‐C, total cholesterol, HbA1c or fasting glucose among the liver fibrosis severity groups (Table 1). This suggests that the pathogenesis of liver fibrosis in patients with stable angina may have a different mechanism to that of liver fibrosis, even though they share similar risk factors.\nIn this study, we found that plasma GDF1 level was independently associated and showed a significant linear trend with liver fibrosis after controlling for all anthropometric variables, liver enzymes, albumin, renal function and inflammation markers (Tables 2 and 3). In linear regression analysis, we further showed that plasma GDF1 level was positively associated with renal function and liver function and condition (such as albumin and platelet levels), but negatively associated with the FIB‐4 index (Table 4). More importantly, the common CAD risk factors such as BMI, blood pressure, Hba1c, LDL, and hs‐CRP were not associated with GDF1. These findings imply that the role of GDF1 in liver fibrosis in patients with stable angina may be different from the traditional shared risk factors between CAD and NAFLD. Further studies are warranted to elucidate this issue.\nAfter classifying the subjects into tertiles according to plasma GDF1 concentration, we found that GDF1 was associated with the FIB‐4 index and APRI, which is used to predict fibrosis and cirrhosis in patients with chronic hepatitis (Figure 1).\n23\n, \n32\n The average AST/ALT ratio in our patients was 1.78, neither <1 or >2 (Figure 1). This revealed that a high FIB‐4 index in our angina patients was not strongly related to the causes of alcoholic fatty liver disease or NAFLD.\n33\n, \n34\n, \n35\n With regard to the general demographics, we found no significant differences in age (69.3 ± 13.4 vs. 69.9 ± 13.9 vs. 68.6 ± 13.3 years, p = 0.769) or alcohol consumption (22.9% vs. 22.2% vs. 34.6%, p = 0.068) among the tertiles of GDF1 (data not shown).\nThere are several limitations to this study. First, we lacked liver echo evaluation reports and liver biopsy proof. This is because it is not reasonable to ask all stable angina patients to undergo such examinations when there is no strong evidence of liver disease. Second, the role of GDF1 in coronary arteriosclerosis remains unknown. Third, our results showed only that there was an association between plasma GDF1 levels and liver fibrosis. Further investigations are needed to investigate whether the mechanism is through the MEK/ERK1/2 signaling pathway or other underlying pathways. Fourth, if the study population had different diseases (e.g., acute coronary syndrome and myocardial infarction), the diverse condition and disease severity of the study population may have impacted the results. To avoid selection bias, we chose individuals with stable angina for this study, thus the results of the present study might not be generalizable to other populations. Finally, further investigations are also needed to investigate whether other hepatokines are also involved and interact with GDF1 in the liver fibrosis process in patients with stable angina.","Our results indicated that in patients with stable angina, low plasma GDF1 was associated with liver fibrosis. Further studies are warranted to investigate the association between plasma GDF1 and the pathogenesis of liver fibrosis.","All authors contributed to this study. W.‐C.H., W.‐H.T., and C.‐C.H. conceived and designed the study. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. provided the methodology. F.‐M.C. performed the formal analysis, and project administration. T.‐H.Y., C.‐C.W., and C.‐C.H. validated the data. T.‐H.Y., C.‐C.W., W.‐C.H., and C.‐P.W. performed the investigation, resources, and data curation. T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. prepared the manuscript. W.‐H.T., T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. reviewed and edited the manuscript. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. performed the visualization. W.‐C.H. and C.‐C.H. performed the supervision and funding acquisition. All authors have read and agreed to the published version of the manuscript.","E‐Da Hospital financially supported this research under Contracts EDAHP109002, EDAHI109002, and EDAHI110001.","The authors have declared that no competing interest exists.","Each patient provided written informed consent before being enrolled into the study."],"string":"[\n \"Cardiovascular disease (CVD) is the leading cause of death worldwide,\\n1\\n and is precipitated by cardiometabolic risk factors including chronic liver disease, diabetes, metabolic syndrome, and chronic renal disease.\\n1\\n, \\n2\\n, \\n3\\n Chronic liver disease is a progressive deterioration of liver function, and the most common causes are hepatitis, alcohol abuse, and nonalcoholic fatty liver disease (NAFLD).\\n4\\n NAFLD is common in patients with CVD, and both diseases share common risks factor such as high cholesterol intake, alcohol consumption, diabetes and other metabolic abnormalities.\\n5\\n In addition, recent studies have reported associations between liver fibrosis, which is the irreversible progression of NAFLD, and coronary artery disease (CAD), cardiac arrhythmias, and impaired heart function.\\n3\\n, \\n6\\n, \\n7\\n Furthermore, previous study also revealed that higher liver fibrosis scores are associated with increased risks of cardiovascular and all‐cause mortality among patients with CAD. Liver fibrosis scores might play a potential role in CAD prognosis prediction.\\n3\\n However, the underlying mechanisms and interactions remain unknown.\\nGrowth differentiation factors (GDFs) are proteins which belong to the transforming growth factor‐β (TGF‐β) superfamily, and 15 GDFs have been identified to date (GDF1 to GDF15).\\n8\\n The TGF‐β superfamily comprised structurally related polypeptides which have been shown to regulate cell differentiation and growth in various adult and embryonic tissues. GDF1 has been shown to be specifically expressed in the nervous systems of adult mice and late‐stage embryos.\\n9\\n, \\n10\\n A previous study showed that GDF1 had favorable effects on vascular endothelial function, sodium excretion, and post‐infarction cardiac remodeling in an animal study and also in patients with type 2 diabetes and normal individuals.\\n11\\n The underlying mechanism is believed to be through the negative regulation of MEK‐ERK1/2 and Smad signaling pathways.\\n11\\n Interestingly, the TGF‐β/SMAD pathway has also been found to play a key role in the development of hepatic fibrosis in subjects with fatty liver disease via suppression of the MEK/ERK1/2 signaling pathway.\\n12\\n, \\n13\\n However, few studies have investigated the relationship between plasma GDF1 level and liver fibrosis. Therefore, the aim of this study was to evaluate the relationship between plasma GDF1 level and liver fibrosis in a cohort of Taiwanese patients with stable angina.\",\n \"[SUBTITLE] Patient population [SUBSECTION] We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\\n14\\n, \\n15\\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\\n16\\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\\nWe enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\\n14\\n, \\n15\\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\\n16\\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\\n[SUBTITLE] Baseline data collection [SUBSECTION] A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\\nA detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\\n[SUBTITLE] Laboratory measurements [SUBSECTION] Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\\n17\\n, \\n18\\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\\n19\\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\\n20\\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\\nFasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\\n17\\n, \\n18\\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\\n19\\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\\n20\\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\\n[SUBTITLE] Measurements of plasma GDF1 and high‐sensitive C‐reactive protein (hs‐CRP) [SUBSECTION] After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\\nAfter obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\\n[SUBTITLE] Definitions [SUBSECTION] We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\\n21\\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\\n22\\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\\n20\\n, \\n23\\n, \\n24\\n\\n\\nWe used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\\n21\\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\\n22\\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\\n20\\n, \\n23\\n, \\n24\\n\\n\\n[SUBTITLE] Statistical analysis [SUBSECTION] Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).\\nContinuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).\",\n \"We enrolled a total of 327 patients with stable angina from July 2012 to December 2021 who visited the cardiovascular clinic at E‐Da Hospital. The inclusion criteria were patients: (1) with stable angina pectoris, defined as effort‐related chest pain without evidence of recent deterioration or rest pain in the previous 6 months, and diagnosed by the cardiologist in charge as previously described;\\n14\\n, \\n15\\n and (2) who underwent a successful percutaneous coronary intervention, defined as <30% residual stenosis in a final angiogram as assessed using quantitative coronary angiography with no occlusion of the large branch (>1 mm) or dissection limiting flow, and Thrombolysis in Myocardial Infarction grade 3.\\n16\\n The exclusion criteria were patients: (1) who used steroids; (2) had inflammatory diseases (such as infection/sepsis), malignancy, liver diseases, and collagen diseases; (3) with a history of psychosis; and (4) with heart valve disease, myocardial infarction, or had undergone heart surgery. In addition, we also excluded patients who were unable or unwilling to provide informed consent. All study patients lived in the same area during the study period, and they were all of Han Chinese ethnicity. Each patient provided written informed consent before being enrolled into the study. This cross‐sectional study was approved by the Human Research Ethics Committee of E‐Da Hospital.\",\n \"A detailed interview was conducted with each patient about their medical and personal history, as well as their demographic characteristics before the coronary angiography examination. We classified smoking status as follows: never smokers, former smokers (those who had stopped smoking for ≥1 year), or current smokers. We grouped the current and former smokers in the analysis and compared them with the never smokers. A trained nurse performed all blood pressure readings with the patients seated using an automated blood pressure monitor (HEM‐907; Omron) after 5 min of rest. Anthropometric parameters including body mass index (BMI) and waist circumference were also recorded.\",\n \"Fasting (8 h) peripheral blood samples were obtained from the antecubital vein before the patients underwent coronary angiography. Levels of high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), total cholesterol, plasma triglycerides, serum glucose, serum albumin, serum creatinine, and complete blood cell count were measured in all patients using a parallel, multi‐channel analyzer (Hitachi 7170A) as described previously.\\n17\\n, \\n18\\n We used high‐performance liquid chromatography (Tosoh Automated Glycohemoglobin Analyzer, HLC‐723G8) to measure hemoglobin A1c (HbA1c). In addition, we measured levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) following the methods of the Japan Society of Clinical Chemistry (“Hitachi” Discrete photometric chemistry analyzer for clinical use, LAbOSPECT 008AS), which is compatible to the methods by the International Federation of Clinical Chemistry. Serum creatinine level was measured using the Jaffe method, and estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation within 3–6 months of admission.\\n19\\n The fibrosis‐4 (FIB‐4) index was calculated as: FIB‐4 = age (years) × AST (IU/L)/platelet count (109/L) × √ALT (IU/L).\\n20\\n The aspartate transaminase/platelet ratio index (APRI) was calculated as: APRI = AST/upper limit of normal of AST/platelet count × 100 (where the upper limit of normal of AST = 40 IU/L).\",\n \"After obtaining overnight fasting blood samples, the plasma was kept at −80°C until assay. We used a commercial ELISA kit to measure concentrations of plasma GDF1 (Cloud‐Clone Corp.). The standard and dilution curves were parallel, and the inter‐assay and intra‐assay coefficients of variation for the assay were <12% and <10%, respectively (both n = 3). Plasma levels of hs‐CRP were measured using an immunochemistry system (Beckman Coulter IMMAGE), which had a detection of 0.2 mg/L. All measurements were made in duplicate during a single experiment.\",\n \"We used the World Health Organization criteria to define type 2 diabetes, as receiving medical therapy for diabetes and/or a history of type 2 diabetes.\\n21\\n Patients with a systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg, or a prescription for antihypertensive drugs were defined as having hypertension. We used the Adult Treatment Panel III criteria\\n22\\n to define hyperlipidemia as: elevated total cholesterol (≥200 mg/dl), and/or elevated LDL‐C (≥130 mg/dl), and/or low HDL‐C (<35 mg/dl in men or < 39 mg/dl in women), and/or elevated triglycerides (≥150 mg/dl), or being treated for a lipid disorder. Liver fibrosis was defined according to FIB‐4 index categories as: minimal fibrosis (FIB‐4 < 1.45), moderate fibrosis (FIB‐4 1.45–3.25), and advanced fibrosis (FIB‐4 > 3.25).\\n20\\n, \\n23\\n, \\n24\\n\\n\",\n \"Continuous data are presented as mean ± SD or median (interquartile range) as appropriate. Between‐group differences were analyzed with one‐way ANOVA for normally distributed variables followed by Tukey's pairwise test. Categorical data are presented as frequency and percentage, and between‐group comparisons were analyzed using the chi‐square test. Because the distributions of serum AST, ALT, triglycerides, creatinine, plasma hs‐CRP, and GDF1 values were skewed, we used logarithmically transformed values in the analysis. The association between GDF1 and liver fibrosis was evaluated using multivariate logistic regression models: (1) GDF1 and age; (2) GDF1, age, and sex; (3) GDF1, age, sex, and BMI; (4) GDF1, age, sex, BMI, and ALT; (5) GDF1, age, sex, BMI, ALT, hemoglobin, and albumin; and (6) GDF1, age, sex, BMI, ALT, hemoglobin, albumin, creatinine, and hs‐CRP. We then classified the concentration of GDF1 into tertiles, and analyzed trends among the tertiles using general linear and logistic regression analyses. Using the highest tertile as the reference, we also estimated the odds ratio (OR) with 95% confidence interval (CI) of liver fibrosis in each tertile.\\nAssociations between plasma GDF1 concentration and other variables were analyzed using simple and multiple linear regression analyses. In addition, the values of FIB‐4 index, AST/ALT ratio and APRI in each GDF1 concentration tertile were analyzed for trends. All tests were 2‐sided, and a p value < 0.05 was considered to be statistically significant. All statistical analyses were performed using JMP version 7.0 for Windows (SAS Institute).\",\n \"[SUBTITLE] Characteristics of the patients by liver fibrosis category [SUBSECTION] Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\\n\\nNote: Data are mean ± SD or frequency (percentage).\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\\nSignificant difference was tested using log‐transformed data.\\nTable 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\\n\\nNote: Data are mean ± SD or frequency (percentage).\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\\nSignificant difference was tested using log‐transformed data.\\n[SUBTITLE] Association between plasma GDF1 and liver fibrosis [SUBSECTION] We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\\n\\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\\n\\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\\nAbbreviation: GDF1, growth/differentiation factor 1.\\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\\nWe found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\\n\\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\\n\\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\\nAbbreviation: GDF1, growth/differentiation factor 1.\\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\\n[SUBTITLE] Associations between plasma GDF1 level and clinical laboratory data [SUBSECTION] Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\\nRegression coefficient adjusted for age and sex.\\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\\nSimple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\\nRegression coefficient adjusted for age and sex.\\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\",\n \"Table 1 shows the biochemical and clinical characteristics of the 327 enrolled patients according to liver fibrosis category. The prevalence rates of minimal liver fibrosis (FIB‐4 < 1.45), moderate liver fibrosis (FIB‐4 1.45–3.25), and advanced liver fibrosis (FIB‐4 > 3.25) were 24.5%, 43.1%, and 32.4%, respectively. The advanced liver fibrosis group were older, had higher levels of AST, blood urea nitrogen, and creatinine, and lower BMI, eGFR, albumin, platelets, hematocrit, and hemoglobin than those with minimal and moderate liver fibrosis. Furthermore, the advanced liver fibrosis group had higher ALT, white blood cell count, and hs‐CRP, and a lower rate of drinking alcohol, waist circumference, and GDF1 than the minimal liver fibrosis group. Moreover, the advanced liver fibrosis group had a lower triglyceride level than the moderate liver fibrosis group. No significant differences were found in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, fasting glucose, HbA1c, total cholesterol, LDL‐C, HDL‐C, Gensini score or number of diseased coronary arteries among the three groups.\\nBaseline characteristics of the study population stratified by severity of liver fibrosis.\\n\\nNote: Data are mean ± SD or frequency (percentage).\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; FIB, fibrosis; GDF1, growth/differentiation factor 1; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; WBC, white blood cell; WHR, waist‐to‐hip ratio.\\nSignificant difference was tested using log‐transformed data.\",\n \"We found a significant association between a lower plasma GDF1 concentration and liver fibrosis, even after controlling for anthropometric factors, ALT, hemoglobin, albumin, creatinine, and hs‐CRP (Table 2). Furthermore, a lower concentration of GDF1 showed a significant linear trend and was independently associated with liver fibrosis, especially when analyzing the concentration both as a continuous variable and by tertile (Tables 2 and 3). In the multiple logistic regression analysis, fully adjusted ORs for liver fibrosis in the second and third tertiles were 2.02 (95% CI, 0.70–6.30) and 5.04 (95% CI, 1.35–25.60), respectively.\\nAssociation of plasma GDF1 with liver fibrosis in fully adjusted models\\n\\nNote: Results of multivariate logistic regression analysis are presented as the odds ratio (OR) of having a liver fibrosis status and decreased plasma GDF1 level.\\nAbbreviations: ALT, alanine aminotransferase; BMI, body mass index; GDF1, growth/differentiation factor 1; Hs‐CRP, high‐sensitivity C‐reactive protein.\\nUnivariate and multivariate analyses of the impact of plasma GDF1 level on liver fibrosis\\n\\nNote: Values shown are cut‐off values of plasma GDF1 levels of all subjects, and odds ratios (ORs) with 95% confidence intervals (CIs).\\nAbbreviation: GDF1, growth/differentiation factor 1.\\nAdjusted for age, sex, body mass index, alanine aminotransferase, hemoglobin, albumin, creatinine, and high‐sensitivity C‐reactive protein.\",\n \"Simple linear regression analysis showed that plasma GDF1 was negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine, and positively associated with platelet count and eGFR. In multiple linear regression analysis, plasma GDF1 level was positively associated with platelet count, and negatively associated with the FIB‐4 index, blood urea nitrogen, and creatinine (Table 4). In addition, in analysis of the subjects by tertile of GDF1 concentration, the FIB‐4 index, APRI, and AST/ALT ratio were significantly associated with GDF1 concentration (p for trend < 0.05, Figure 1).\\nLinear regression analysis of variables associated with plasma log‐GDF1 levels in the study subjects.\\nAbbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BP, blood.pressure; GDF1, growth/differentiation factor 1; GFR, glomerular filtration rate; HDL, high‐density lipoprotein; Hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein.\\nRegression coefficient adjusted for age and sex.\\nClassification of subjects into tertiles according to growth/differentiation factor 1 (GDF1) concentration revealed that the fibrosis‐4 index (A), APRI (B), and AST/ALT ratio (C) were significantly associated with GDF1 concentration (p for trend < 0.05). Bars represent the mean ± SD. ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase.\",\n \"According to a previous clinical study, fatty liver disease is common in patients with CAD, and it can progress to liver fibrosis resulting in higher rates of all‐cause and cardiovascular mortality.\\n3\\n Similarly, in the current study, we found that among the 327 stable angina patients, more than 75% had moderate to advanced liver fibrosis (Table 1). In the past decade, several studies have evaluated the relationship between fatty liver disease and CAD. Although some chronic liver diseases such as NAFLD share similar risk factors with CAD (such as hyperglycemia, insulin resistance, hypertension, inflammation, dyslipidemia, hyperuricemia, hypoadiponectinemia, renal failure, and obesity),\\n25\\n, \\n26\\n, \\n27\\n the underlying molecular relationship between chronic liver disease and CAD is unknown. Clinically, a meta‐analysis found that NAFLD was associated with the severity of atherosclerosis, including coronary calcification levels, carotid intima‐media thickness, arterial stiffness and endothelial dysfunction.\\n28\\n Furthermore, the review by Meex et al. in 2017 also showed that many hepatokines such as fetuin A, fetuin B, retinol‐binding protein 4, and selenoprotein P are involved in both fatty liver disease and insulin resistance, which in turn influences the process of atherosclerosis.\\n29\\n However, no previous report has described the underlying molecular pathogenesis between chronic liver disease and CAD. To the best of our knowledge, this is the first study to show that GDF1, which is involved in both post‐infarction cardiac remodeling and the development of fatty liver disease, was associated with liver fibrosis in patients with stable angina.\\nThe cytokine system is very complex. Maintaining a balanced healthy environment for all human cells is difficult, and a factor which is beneficial for certain circumstances may be harmful for another. GDF1 is a member of the TGF‐β superfamily and is known to be a protective mediator against the development of post‐infarction cardiac remodeling via negative regulation of the Smad signaling and MEK‐ERK1/2 pathways in the heart.\\n11\\n, \\n30\\n, \\n31\\n A suppressed MEK/ERK1/2 signaling pathway has been associated with fatty liver disease, liver fibrosis, and poor hepatoma differentiation in recent studies.\\n12\\n, \\n13\\n In this study, a lower plasma GDF1 concentration was significantly independently associated with and showed a significant linear trend with liver fibrosis (Tables 2 and 3). It is reasonable that a decrease in GDF1 could enhance the progress of liver fibrosis. In addition, the Gensini score and number of diseased coronary arteries, which was used to evaluate CAD severity, did not show a significant difference among tertiles of GDF1 (data not shown). The role of GDF1 post infarction is known, however it remains unclear in CAD.\\nIn analysis of the associations between common risk factors for liver fibrosis, we found no significant differences in hypertension, male sex, hyperlipidemia, current smoking, diabetes mellitus, DBP, SBP, LDL‐C, HDL‐C, total cholesterol, HbA1c or fasting glucose among the liver fibrosis severity groups (Table 1). This suggests that the pathogenesis of liver fibrosis in patients with stable angina may have a different mechanism to that of liver fibrosis, even though they share similar risk factors.\\nIn this study, we found that plasma GDF1 level was independently associated and showed a significant linear trend with liver fibrosis after controlling for all anthropometric variables, liver enzymes, albumin, renal function and inflammation markers (Tables 2 and 3). In linear regression analysis, we further showed that plasma GDF1 level was positively associated with renal function and liver function and condition (such as albumin and platelet levels), but negatively associated with the FIB‐4 index (Table 4). More importantly, the common CAD risk factors such as BMI, blood pressure, Hba1c, LDL, and hs‐CRP were not associated with GDF1. These findings imply that the role of GDF1 in liver fibrosis in patients with stable angina may be different from the traditional shared risk factors between CAD and NAFLD. Further studies are warranted to elucidate this issue.\\nAfter classifying the subjects into tertiles according to plasma GDF1 concentration, we found that GDF1 was associated with the FIB‐4 index and APRI, which is used to predict fibrosis and cirrhosis in patients with chronic hepatitis (Figure 1).\\n23\\n, \\n32\\n The average AST/ALT ratio in our patients was 1.78, neither <1 or >2 (Figure 1). This revealed that a high FIB‐4 index in our angina patients was not strongly related to the causes of alcoholic fatty liver disease or NAFLD.\\n33\\n, \\n34\\n, \\n35\\n With regard to the general demographics, we found no significant differences in age (69.3 ± 13.4 vs. 69.9 ± 13.9 vs. 68.6 ± 13.3 years, p = 0.769) or alcohol consumption (22.9% vs. 22.2% vs. 34.6%, p = 0.068) among the tertiles of GDF1 (data not shown).\\nThere are several limitations to this study. First, we lacked liver echo evaluation reports and liver biopsy proof. This is because it is not reasonable to ask all stable angina patients to undergo such examinations when there is no strong evidence of liver disease. Second, the role of GDF1 in coronary arteriosclerosis remains unknown. Third, our results showed only that there was an association between plasma GDF1 levels and liver fibrosis. Further investigations are needed to investigate whether the mechanism is through the MEK/ERK1/2 signaling pathway or other underlying pathways. Fourth, if the study population had different diseases (e.g., acute coronary syndrome and myocardial infarction), the diverse condition and disease severity of the study population may have impacted the results. To avoid selection bias, we chose individuals with stable angina for this study, thus the results of the present study might not be generalizable to other populations. Finally, further investigations are also needed to investigate whether other hepatokines are also involved and interact with GDF1 in the liver fibrosis process in patients with stable angina.\",\n \"Our results indicated that in patients with stable angina, low plasma GDF1 was associated with liver fibrosis. Further studies are warranted to investigate the association between plasma GDF1 and the pathogenesis of liver fibrosis.\",\n \"All authors contributed to this study. W.‐C.H., W.‐H.T., and C.‐C.H. conceived and designed the study. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. provided the methodology. F.‐M.C. performed the formal analysis, and project administration. T.‐H.Y., C.‐C.W., and C.‐C.H. validated the data. T.‐H.Y., C.‐C.W., W.‐C.H., and C.‐P.W. performed the investigation, resources, and data curation. T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. prepared the manuscript. W.‐H.T., T.‐H.Y., W.‐C.H., C.‐C.W., C.‐P.W., Y.‐C.L., and C.‐T.W. reviewed and edited the manuscript. W.‐C.H., W.‐H.T., Y.‐J.L., and C.‐C.H. performed the visualization. W.‐C.H. and C.‐C.H. performed the supervision and funding acquisition. All authors have read and agreed to the published version of the manuscript.\",\n \"E‐Da Hospital financially supported this research under Contracts EDAHP109002, EDAHI109002, and EDAHI110001.\",\n \"The authors have declared that no competing interest exists.\",\n \"Each patient provided written informed consent before being enrolled into the study.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"materials-and-methods",null,null,null,null,null,null,"results",null,null,null,"discussion","conclusions",null,null,"COI-statement",null],"string":"[\n null,\n \"materials-and-methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n \"discussion\",\n \"conclusions\",\n null,\n null,\n \"COI-statement\",\n null\n]"},"keywords":{"kind":"list like","value":["growth/differentiation factor 1","liver fibrosis","MEK‐ERK1/2","Smad signaling","stable angina"],"string":"[\n \"growth/differentiation factor 1\",\n \"liver fibrosis\",\n \"MEK‐ERK1/2\",\n \"Smad signaling\",\n \"stable angina\"\n]"}}},{"rowIdx":385,"cells":{"title":{"kind":"string","value":"Clinical features of patients with talaromycosis marneffei and microbiological characteristics of the causative strains."},"pmid":{"kind":"string","value":"36268985"},"background_abstract":{"kind":"string","value":"Talaromyces marneffei (T. marneffei) is a temperature-dependent dimorphic fungus that is mainly prevalent in Southeast Asia and South China and often causes disseminated life-threatening infections. This study aimed to investigate the clinical features and improve the early diagnosis of talaromycosis marneffei in nonendemic areas."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We retrospectively analyzed the medical records of six cases of T. marneffei infection. We describe the clinical manifestations, laboratory tests, and imaging manifestations of the six patients."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Talaromyces marneffei infection was confirmed by sputum culture, blood culture, tissue biopsy, and metagenomic next-generation sequencing (mNGS). In this study, there were five disseminated-type patients and two HIV patients. One patient died within 24 h, and the others demonstrated considerable improvement after definitive diagnosis."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Due to the lack of significant clinical presentations of talaromycosis marneffei, many cases may be easily misdiagnosed in nonendemic areas. It is particularly important to analyze the imaging manifestations and laboratory findings of infected patients. With the rapid development of molecular biology, mNGS may be a rapid and effective diagnostic method."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","HIV Infections","Retrospective Studies","Mycoses","China","Antifungal Agents"],"string":"[\n \"Humans\",\n \"HIV Infections\",\n \"Retrospective Studies\",\n \"Mycoses\",\n \"China\",\n \"Antifungal Agents\"\n]"},"pmcid":{"kind":"string","value":"9701894"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"string","value":"METHODS"},"methods_text":{"kind":"string","value":"[SUBTITLE] Clinical data [SUBSECTION] This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\nThis was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\n[SUBTITLE] Laboratory tests [SUBSECTION] Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\nRoutine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\n[SUBTITLE] Identification methods of T. marneffei\n [SUBSECTION] There were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThere were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results."},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Clinical manifestation [SUBSECTION] Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\nSix patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\n[SUBTITLE] Laboratory tests [SUBSECTION] C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\nC‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\n[SUBTITLE] Imaging and other examinations [SUBSECTION] Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\nAmong six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\n[SUBTITLE] Diagnosis of patients with talaromycosis marneffei [SUBSECTION] Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nSix cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Clinical data","Laboratory tests","Identification methods of T. marneffei\n","Pathogen culture","Microscopy observation","Metagenomic next‐generation sequencing","Clinical manifestation","Laboratory tests","Imaging and other examinations","Diagnosis of patients with talaromycosis marneffei","Culture and microscopy observation","Metagenomic next‐generation sequencing","TREATMENT AND PROGNOSIS","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION"],"string":"[\n \"INTRODUCTION\",\n \"Clinical data\",\n \"Laboratory tests\",\n \"Identification methods of T. marneffei\\n\",\n \"Pathogen culture\",\n \"Microscopy observation\",\n \"Metagenomic next‐generation sequencing\",\n \"Clinical manifestation\",\n \"Laboratory tests\",\n \"Imaging and other examinations\",\n \"Diagnosis of patients with talaromycosis marneffei\",\n \"Culture and microscopy observation\",\n \"Metagenomic next‐generation sequencing\",\n \"TREATMENT AND PROGNOSIS\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\"\n]"},"other_sections_texts":{"kind":"list like","value":["\nTalaromyces marneffei is a rare pathogenic thermally dimorphic fungus that grows as mold at 25°C and as yeast at 37°C.\n1\n The resistance of the host to invasion by T. marneffei may be based on cellular immunity, with the T‐cell‐mediated Thl‐type response pattern playing an important role in the defense of the host against fungal infection.\n2\n It mainly invades the human monocyte–macrophage system and causes infections in patients with immune dysfunction (e.g., immunosuppressed, immunocompromised, or immunodeficient patients). The disease is predominant in Guangxi, Hong Kong, and Taiwan but rare in the city of Ningbo in Zhejiang Province.\n3\n, \n4\n It has become the most common opportunistic fungal infection in Southeast Asia (especially in AIDS patients).\n5\n However, in recent years, there have been increasing reports of cases among non‐HIV‐infected patients.\n6\n In this article, we describe the clinical and laboratory features of six patients.","This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.","Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.","There were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.","Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.","\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.","The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.","Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.","C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.","Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).","Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.","The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).","Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.","Amphotericin B, itraconazole, and voriconazole are routinely utilized for talaromycosis marneffei in clinical treatment. In terms of clinical response and fungicidal action, amphotericin B was superior to itraconazole in the initial treatment.\n7\n Voriconazole was reported to be effective for disseminated T. marneffei infection.\n8\n Thus, most of the cases in this study were treated with intravenous amphotericin B at a dose of 0.6 mg/kg/day for 2 weeks, followed by intravenous voriconazole at a dose of 200 mg/day for maintenance therapy. One mild case was treated with intravenous voriconazole at a dose of 200 mg q12 h for the first 24 h, followed by oral therapy at 200 mg twice daily. To reduce the incidence of adverse effects, liver function, renal function, and body electrolytes were monitored closely and regularly during treatment.\nDuring the treatment period, adverse drug reactions to amphotericin B were observed. Two cases showed elevated creatinine levels. These patients were treated with lower doses of amphotericin B, or the drug was discontinued and replaced with voriconazole therapy. Only one patient did not receive antifungal medication and died within 24 h of admission due to deterioration of the disease. During the 6‐month follow‐up, most of the cases in this study showed improvement after treatment and had a good prognosis. One case is still under treatment and follow‐up.","All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Lei Peng and Xing‐bei Weng. The first draft of the article was written by Lei Peng, and all authors commented on previous versions of the article. All authors read and approved the final article.","This work was supported by the Ningbo Natural Science Foundation (2019A610381) and the Ningbo Public Welfare Foundation (2019C50087)."],"string":"[\n \"\\nTalaromyces marneffei is a rare pathogenic thermally dimorphic fungus that grows as mold at 25°C and as yeast at 37°C.\\n1\\n The resistance of the host to invasion by T. marneffei may be based on cellular immunity, with the T‐cell‐mediated Thl‐type response pattern playing an important role in the defense of the host against fungal infection.\\n2\\n It mainly invades the human monocyte–macrophage system and causes infections in patients with immune dysfunction (e.g., immunosuppressed, immunocompromised, or immunodeficient patients). The disease is predominant in Guangxi, Hong Kong, and Taiwan but rare in the city of Ningbo in Zhejiang Province.\\n3\\n, \\n4\\n It has become the most common opportunistic fungal infection in Southeast Asia (especially in AIDS patients).\\n5\\n However, in recent years, there have been increasing reports of cases among non‐HIV‐infected patients.\\n6\\n In this article, we describe the clinical and laboratory features of six patients.\",\n \"This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\",\n \"Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\",\n \"There were three methods used for pathogen examination.\\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\n[SUBTITLE] Microscopy observation [SUBSECTION] \\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\",\n \"Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\",\n \"\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\",\n \"The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\",\n \"Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\\nClinical features of patients with talaromycosis marneffei\\n\\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\",\n \"C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\\n\\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\",\n \"Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\",\n \"Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\",\n \"The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\",\n \"Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\",\n \"Amphotericin B, itraconazole, and voriconazole are routinely utilized for talaromycosis marneffei in clinical treatment. In terms of clinical response and fungicidal action, amphotericin B was superior to itraconazole in the initial treatment.\\n7\\n Voriconazole was reported to be effective for disseminated T. marneffei infection.\\n8\\n Thus, most of the cases in this study were treated with intravenous amphotericin B at a dose of 0.6 mg/kg/day for 2 weeks, followed by intravenous voriconazole at a dose of 200 mg/day for maintenance therapy. One mild case was treated with intravenous voriconazole at a dose of 200 mg q12 h for the first 24 h, followed by oral therapy at 200 mg twice daily. To reduce the incidence of adverse effects, liver function, renal function, and body electrolytes were monitored closely and regularly during treatment.\\nDuring the treatment period, adverse drug reactions to amphotericin B were observed. Two cases showed elevated creatinine levels. These patients were treated with lower doses of amphotericin B, or the drug was discontinued and replaced with voriconazole therapy. Only one patient did not receive antifungal medication and died within 24 h of admission due to deterioration of the disease. During the 6‐month follow‐up, most of the cases in this study showed improvement after treatment and had a good prognosis. One case is still under treatment and follow‐up.\",\n \"All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Lei Peng and Xing‐bei Weng. The first draft of the article was written by Lei Peng, and all authors commented on previous versions of the article. All authors read and approved the final article.\",\n \"This work was supported by the Ningbo Natural Science Foundation (2019A610381) and the Ningbo Public Welfare Foundation (2019C50087).\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","METHODS","Clinical data","Laboratory tests","Identification methods of T. marneffei\n","Pathogen culture","Microscopy observation","Metagenomic next‐generation sequencing","RESULTS","Clinical manifestation","Laboratory tests","Imaging and other examinations","Diagnosis of patients with talaromycosis marneffei","Culture and microscopy observation","Metagenomic next‐generation sequencing","TREATMENT AND PROGNOSIS","DISCUSSION","AUTHOR CONTRIBUTIONS","FUNDING INFORMATION","CONFLICT OF INTEREST"],"string":"[\n \"INTRODUCTION\",\n \"METHODS\",\n \"Clinical data\",\n \"Laboratory tests\",\n \"Identification methods of T. marneffei\\n\",\n \"Pathogen culture\",\n \"Microscopy observation\",\n \"Metagenomic next‐generation sequencing\",\n \"RESULTS\",\n \"Clinical manifestation\",\n \"Laboratory tests\",\n \"Imaging and other examinations\",\n \"Diagnosis of patients with talaromycosis marneffei\",\n \"Culture and microscopy observation\",\n \"Metagenomic next‐generation sequencing\",\n \"TREATMENT AND PROGNOSIS\",\n \"DISCUSSION\",\n \"AUTHOR CONTRIBUTIONS\",\n \"FUNDING INFORMATION\",\n \"CONFLICT OF INTEREST\"\n]"},"all_sections_texts":{"kind":"list like","value":["\nTalaromyces marneffei is a rare pathogenic thermally dimorphic fungus that grows as mold at 25°C and as yeast at 37°C.\n1\n The resistance of the host to invasion by T. marneffei may be based on cellular immunity, with the T‐cell‐mediated Thl‐type response pattern playing an important role in the defense of the host against fungal infection.\n2\n It mainly invades the human monocyte–macrophage system and causes infections in patients with immune dysfunction (e.g., immunosuppressed, immunocompromised, or immunodeficient patients). The disease is predominant in Guangxi, Hong Kong, and Taiwan but rare in the city of Ningbo in Zhejiang Province.\n3\n, \n4\n It has become the most common opportunistic fungal infection in Southeast Asia (especially in AIDS patients).\n5\n However, in recent years, there have been increasing reports of cases among non‐HIV‐infected patients.\n6\n In this article, we describe the clinical and laboratory features of six patients.","[SUBTITLE] Clinical data [SUBSECTION] This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\nThis was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\n[SUBTITLE] Laboratory tests [SUBSECTION] Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\nRoutine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\n[SUBTITLE] Identification methods of T. marneffei\n [SUBSECTION] There were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThere were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.","This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.","Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.","There were three methods used for pathogen examination.\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\n[SUBTITLE] Microscopy observation [SUBSECTION] \nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.","Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.","\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.","The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.","[SUBTITLE] Clinical manifestation [SUBSECTION] Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\nSix patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\n[SUBTITLE] Laboratory tests [SUBSECTION] C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\nC‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\n[SUBTITLE] Imaging and other examinations [SUBSECTION] Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\nAmong six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\n[SUBTITLE] Diagnosis of patients with talaromycosis marneffei [SUBSECTION] Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nSix cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.","Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\nClinical features of patients with talaromycosis marneffei\n\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.","C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\n\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.","Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).","Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.","The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).","Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\nMicroorganism detected by mNGS of two patients\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.","Amphotericin B, itraconazole, and voriconazole are routinely utilized for talaromycosis marneffei in clinical treatment. In terms of clinical response and fungicidal action, amphotericin B was superior to itraconazole in the initial treatment.\n7\n Voriconazole was reported to be effective for disseminated T. marneffei infection.\n8\n Thus, most of the cases in this study were treated with intravenous amphotericin B at a dose of 0.6 mg/kg/day for 2 weeks, followed by intravenous voriconazole at a dose of 200 mg/day for maintenance therapy. One mild case was treated with intravenous voriconazole at a dose of 200 mg q12 h for the first 24 h, followed by oral therapy at 200 mg twice daily. To reduce the incidence of adverse effects, liver function, renal function, and body electrolytes were monitored closely and regularly during treatment.\nDuring the treatment period, adverse drug reactions to amphotericin B were observed. Two cases showed elevated creatinine levels. These patients were treated with lower doses of amphotericin B, or the drug was discontinued and replaced with voriconazole therapy. Only one patient did not receive antifungal medication and died within 24 h of admission due to deterioration of the disease. During the 6‐month follow‐up, most of the cases in this study showed improvement after treatment and had a good prognosis. One case is still under treatment and follow‐up.","\nTalaromyces marneffei is a species of penicillium isolated from the liver of bamboo rats that died unnaturally in Vietnam in 1956. In 1959, the fungus was named Penicillium marneffei in honor of Hubert Marneffe, the director of the Pasteur Institute in Indochina.\n9\n In 2011, based on phylogenetic and phenotypic analysis, the name was changed to T. marneffei.\n10\n Bamboo rats are currently considered natural hosts or carriers of this pathogen.\n11\n Regarding the patients' epidemiological history, it was found that no patients had a history of residence in endemic areas or exposure to bamboo rats. The original route of infection is currently unknown, but it could be through skin lesions or conidia inhalation. It has been documented that T. marneffei enters the blood system by invading the reticuloendothelial system and then spreads to other organs through the lymphatic system and blood circulation.\n12\n, \n13\n\nT. marneffei mainly invades organs and sites related to the monocyte–macrophage reticuloendothelial system. T. marneffei infections can be divided into disseminated and focal types; focal type infections are often confined to the site of invasion, and the clinical manifestations are dominated by the symptoms of the original disease. The disseminated type often involves multiple tissues and organs. Common clinical manifestations of disseminated T. marneffei include fever, respiratory signs, anemia, weight loss, skin lesions, lymphadenopathy, and hepatosplenomegaly.\n3\n In this study, five patients exhibited disseminated T. marneffei infections. These patients presented with fever, cough, lymph node enlargement, anemia, and gastrointestinal symptoms, all of which were nonspecific and led to a misdiagnosis of tuberculosis or other fungal infections. Skin lesions are a clinical feature of disseminated T. marneffei, and characteristic lesions are papules with central necrosis.\n9\n Two patients in our cohort presented with atypical skin lesions, such as subcutaneous nodules or abscesses. In three cases, the lungs were affected by fungal invasion first. In addition, the possibility of gastrointestinal infection was not excluded. Although lymphoid tissue is extensive throughout the digestive tract, talaromycosis marneffei infection of the intestine is uncommon.\n13\n The atypical and nonspecific clinical presentation of fungal infections in T. marneffei often adds to the difficulty of early diagnosis, resulting in diagnostic delay and increased mortality.\nT lymphocyte cells are critical for controlling T. marneffei infection.\n14\n Some experiments have indicated that CD4+ T cells mediate significant pulmonary inflammatory infiltration in mice by recruiting macrophages and granulocytes. CD4 deficiency may lead to the loss of Th2 anti‐inflammatory cytokines (IL‐4, IL‐5, and IL‐10).\n15\n In the innate immune system, activated macrophages can activate some proinflammatory cytokines and chemokines.\n16\n Myeloperoxidase‐dependent neutrophils also have fungicidal activity through the actions of reactive oxygen species.\n17\n In HIV‐positive individuals, T. marneffei‐activated monocyte‐derived dendritic cells (MDDCs) may further accelerate immunosuppression by promoting HIV‐1 transfection of CD4+ T cells.\n18\n Overall, the inflammatory and regulatory responses to a fungal infection are significantly modulated by the innate and acquired immune systems. Lymphocyte subpopulation analysis should be performed in clinical work, especially in immunosuppressed or HIV‐positive patients.\nThe pathogenesis of T. marneffei involves the ability of the fungus to avoid being killed by macrophages and replicate inside them.\n19\n\nT. marneffei can produce melanin or melanin‐like compounds to enhance resistance to phagocytosis by macrophages.\n20\n After phagocytosis by macrophages, T. marneffei (as an intracellular yeast form) induces the production of superoxide dismutase and catalase‐peroxidase in response to oxidative stress.\n19\n Other factors may also be associated with the pathogenicity of T. marneffei, including heat shock proteins (Hsp70 and Hsp30),\n19\n secreted galactomannan protein Mp1p\n21\n and the expression of isocitrate lyase.\n22\n Subsequently, conidial replication and establishment of infection mediate disseminated infections.\nThe incidence of infection in HIV‐negative patients is increasing. For example, in 2012, Yongxuan Hu reviewed 668 cases of talaromycosis marneffei in mainland China from January 1984 to December 2009, and there were 82 (12.3%) HIV‐negative infected patients.\n3\n In 2015, Ye Qiu retrospectively analyzed 109 patients admitted to the First Affiliated Hospital of Guangxi Medical University for talaromycosis marneffei from January 1, 2003, to August 1, 2014, 39.45% of whom were HIV‐negative.\n23\n In HIV‐negative individuals, infection factors are most likely attributed to underlying conditions, including immunodeficiency due to anti‐IFN‐γ autoantibodies, immunosuppressive therapy, malignancy, diabetes mellitus, and organ transplantation.\n24\n Anti‐IFN‐γ autoantibodies can inhibit STAT1 phosphorylation and IL‐12 production, increasing the risk of fungal infection.\n25\n Other predisposing factors may be related to lymphocyte depletion and impairment of lymphocyte proliferation. Interestingly, one of our cases was an immunocompetent patient, and the infection factor was most likely attributed to the patient's diabetes (Table 2).\nIn terms of laboratory tests, all patients had elevated serum CRP levels and erythrocyte sedimentation rates (Table 1). Some patients had liver damage, characterized by elevated aspartate aminotransferase (AST), especially severe patients. HIV‐positive patients were more likely to have leukopenia, low platelet counts, elevation of alanine transaminase, and positive blood cultures. In addition, the chest imaging manifestations of patients with T. marneffei infection were diverse, including multiple patches or large consolidations, ground‐glass opacities, pleural effusion, and hilar and mediastinal lymph node enlargement (Table 1). Infectious diseases such as tuberculosis, Aspergillus pneumonia, lymphoma, lung cancer, and others should be ruled out.\n26\n\n\nThe traditional methods for diagnosis of talaromycosis marneffei include culture of the pathogen and histopathological examination.\n1\n However, the culture positivity rate was not high in patients with early infection. These sample culture results were presented as repeated negative, including a blood culture of case 3 and a stool culture of case 4. Among the six cases, two were diagnosed with talaromycosis marneffei by mNGS. One case was diagnosed with talaromycosis marneffei by examination of a peripheral blood smear (Figure 4). Yeast‐like organisms may be seen on the blood smear, especially in patients with heavy fungemia.\n27\n Blood culture was positive in only two patients in our study (Table 2). Whereas it usually takes 3–10 days for a patient's clinical specimen to be cultured for T. marneffei, the time from admission to diagnosis was long for most patients in our study. During this period, some patients were misdiagnosed with bacterial pneumonia or other diseases. Moreover, they received improper treatment. When none of their conditions improved significantly, physicians began to suspect the initial diagnosis and considered mNGS. Other techniques, such as novel Mp1p enzyme immunoassay and the beta‐D‐glucan test, have a certain role in the auxiliary diagnosis of talaromycosis marneffei.\n28\n, \n29\n However, these approaches have the drawbacks of extensive processes and limited sensitivity. Several real‐time PCR techniques to identify T. marneffei have been developed in response to the demand for faster and more sensitive diagnostics, although these are limited to primer and probe design.\n30\n, \n31\n Under such circumstances, the implementation of NGS in the clinical field might have some undeniable advantages.\nmNGS, a rapid and effective diagnostic approach, can obtain information on microbial species and abundance in samples by high‐throughput sequencing of nucleic acids and bioinformatics analysis.\n32\n It not only comprehensively identifies pathogens but also offers a significantly reduced time requirement for pathogen identification. In this study, clinicians chose mNGS in the following situations: (1) difficulty identifying fungal pathogens despite repeated cultures and (2) a need for early and precise diagnosis in patients with severe pathogenic infections. Multiple studies have reported the use of mNGS in clinical fungal diagnosis, which also reveals the potential merits of this technique in rapid etiological diagnosis.\n32\n, \n33\n, \n34\n, \n35\n, \n36\n However, there are some shortcomings in the widespread application of mNGS, such as high cost and a lack of recognized standards in this area.\nTalaromycosis marneffei is an uncommon disease that causes high mortality rates. A previous study showed that HIV infection, CD4/CD8 < 0.5, a reduced percentage of CD4+ T cells and late diagnosis were potential risk factors for poor prognosis.\n37\n Late diagnosis is the leading cause of death in T. marneffei‐infected individuals.\n38\n It is clear that early and rapid diagnosis is particularly important. Therefore, clinicians should consider talaromycosis marneffei if a patient has suffered recurrent fever; multiple patches, nodules, or masses in the lungs; gastrointestinal symptoms; and characteristic skin lesions, especially if anti‐inflammatory treatment is ineffective. Pathogen cultivation and other relevant laboratory tests are essential. In addition, we recommend the early use of mNGS to reduce mortality and prolong survival time. There are some limitations of this study. First, it is a retrospective cohort study, and the limited number of cases makes it difficult to explore the differences in clinical manifestations between HIV‐positive and HIV‐negative populations. Second, there is a lack of comprehensive analysis of the infection mechanism of T. marneffei in nonendemic areas.\nIn conclusion, we analyzed six cases of talaromycosis marneffei in nonendemic areas in the hope of promoting awareness and familiarity with this rare disease among more clinicians in nonendemic areas and improving the prognosis of patients.","All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Lei Peng and Xing‐bei Weng. The first draft of the article was written by Lei Peng, and all authors commented on previous versions of the article. All authors read and approved the final article.","This work was supported by the Ningbo Natural Science Foundation (2019A610381) and the Ningbo Public Welfare Foundation (2019C50087).","The authors declare no competing interests."],"string":"[\n \"\\nTalaromyces marneffei is a rare pathogenic thermally dimorphic fungus that grows as mold at 25°C and as yeast at 37°C.\\n1\\n The resistance of the host to invasion by T. marneffei may be based on cellular immunity, with the T‐cell‐mediated Thl‐type response pattern playing an important role in the defense of the host against fungal infection.\\n2\\n It mainly invades the human monocyte–macrophage system and causes infections in patients with immune dysfunction (e.g., immunosuppressed, immunocompromised, or immunodeficient patients). The disease is predominant in Guangxi, Hong Kong, and Taiwan but rare in the city of Ningbo in Zhejiang Province.\\n3\\n, \\n4\\n It has become the most common opportunistic fungal infection in Southeast Asia (especially in AIDS patients).\\n5\\n However, in recent years, there have been increasing reports of cases among non‐HIV‐infected patients.\\n6\\n In this article, we describe the clinical and laboratory features of six patients.\",\n \"[SUBTITLE] Clinical data [SUBSECTION] This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\\nThis was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\\n[SUBTITLE] Laboratory tests [SUBSECTION] Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\\nRoutine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\\n[SUBTITLE] Identification methods of T. marneffei\\n [SUBSECTION] There were three methods used for pathogen examination.\\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\n[SUBTITLE] Microscopy observation [SUBSECTION] \\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\\nThere were three methods used for pathogen examination.\\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\n[SUBTITLE] Microscopy observation [SUBSECTION] \\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\",\n \"This was a retrospective study that used data collected from medical records. Six patients were diagnosed and treated at the Ningbo First Hospital, and the patient's inpatient medical records were reviewed. Related data included demographic information (age and sex), clinical features (with or without underlying disease), laboratory data, imaging findings, treatment, and prognoses. The clinical data of the patients are shown in Table 1.\\nClinical features and laboratory findings of six patients with Talaromyces marneffei infection\\nAbbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; CRP, C‐reactive protein; ESR, erythrocyte sedimentation rate.\",\n \"Routine blood examination, HIV antibody, C‐reactive protein (CRP), blood coagulation, erythrocyte sedimentation rate, and blood biochemistry tests were performed with standardized testing equipment and kits.\",\n \"There were three methods used for pathogen examination.\\n[SUBTITLE] Pathogen culture [SUBSECTION] Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\nCultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\\n[SUBTITLE] Microscopy observation [SUBSECTION] \\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\\nThe metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\",\n \"Cultures of clinical specimens, including blood, sputum, and bronchoalveolar lavage fluid, were established on Sabouraud agar medium at 25°C and 37°C. Identification was based upon the morphology of the colonies. At 25°C, T. marneffei grew as a mold and created a soluble red pigment that diffused into the agar. At 37°C, it developed as a yeast.\",\n \"\\nTalaromyces marneffei was identified by cytology and histopathology from clinical specimens by hexamine silver staining and Wright–Giemsa staining.\",\n \"The metagenomic next‐generation sequencing (mNGS) method is based on the Illumina sequencing platform and polymerase chain reaction (PCR) free library building technology. In this study, clinical specimens were sent to specialized testing institutions for different sample processing procedures. This approach mainly includes specimen pretreatment, nucleic acid extraction, library preparation, online sequencing, database matching, and interpretation of results.\",\n \"[SUBTITLE] Clinical manifestation [SUBSECTION] Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\\nClinical features of patients with talaromycosis marneffei\\n\\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\\nSix patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\\nClinical features of patients with talaromycosis marneffei\\n\\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\\n[SUBTITLE] Laboratory tests [SUBSECTION] C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\\n\\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\\nC‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\\n\\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\\n[SUBTITLE] Imaging and other examinations [SUBSECTION] Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\\nAmong six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\\n[SUBTITLE] Diagnosis of patients with talaromycosis marneffei [SUBSECTION] Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\\nSix cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\",\n \"Six patients were diagnosed and treated at the Ningbo First Hospital, all of whom had lived in Ningbo for a long time. There were two female and four male patients, and all had no epidemiological history and no history of exposure to bamboo rats. The median age of the patients was 60 years.\\nThe most common clinical symptoms of patients infected with T. marneffei were fever and anemia, followed by fungemia, malaise, respiratory symptoms, and weight loss (Table 2). The primary presentation for one patient was gastrointestinal symptoms (such as abdominal pain, diarrhea, and bloody stools). Five patients exhibited disseminated T. marneffei infections; three patients had primary lesions in the lungs, and two patients presented with subcutaneous nodules or abscesses (Table 1). All had various underlying diseases, including chronic obstructive pulmonary disease, multiple myeloma, diabetes, and HIV. Notably, four cases were initially misdiagnosed as pneumonia or pulmonary tuberculosis.\\nClinical features of patients with talaromycosis marneffei\\n\\nNote: Cured: the resolution of all symptoms and mycological eradication; Recovered: improvement in most signs and symptoms; Relapsed: no clinical improvement, clinical condition deterioration, or death.\\nAbbreviations: BALF, bronchoalveolar lavage fluid; COPD, chronic obstructive pulmonary disease; F, female; M, male; TM, talaromycosis marneffei.\",\n \"C‐reactive protein (CRP) and erythrocyte sedimentation rate were significantly elevated in six patients, and the white blood cell count was increased in four patients and decreased in one patient. In addition, five patients had varying degrees of anemia, all below 100 g/L; four patients had varying degrees of decreased albumin, with the lowest value being 22.4 g/L; and three patients were tested for lymphocyte subpopulation, which reflected low immune function in two patients and normal immune function in the other (Table 3). Two patients tested positive for HIV antibody (primary screening) and were diagnosed with HIV infection by the Centers for Disease Control and Prevention. Three patients had no significant abnormalities in liver and kidney function.\\nAnalysis of lymphocyte subsets in three patients with talaromycosis marneffei\\n\\nNote: Normal ranges: CD3: 58.4%–81.56%; CD4: 31%–60%; CD8: 13%–41%; CD4/CD8: 0.8–4.2; NK: 14%–40%; CD19: 5%–25%.\",\n \"Among six patients, chest imaging showed abnormal lesions in the lungs of five patients for the first time (Table 4). There were multiple enlarged lymph nodes in the mediastinum and pulmonary hila in two patients (Figure 1A), multiple high‐density plaques and strips in two patients (Figure 1B), multiple patchy ground‐glass opacities in one patient (Figure 1C), and bilateral pleural effusion in three patients (Figure 1D).\\nHigh‐resolution computed tomography among six patients infected with Talaromyces marneffei.\\nChest CT images among patients infected with Talaromyces marneffei. (A) The blue arrow indicates enlarged mediastinal lymph nodes (mediastinal window). (B) The blue arrow indicates high‐density plaque and strip. (C) Blue arrows indicate multiple patchy ground‐glass opacities. (D) Blue arrows indicate bilateral pleural effusion.\\nCase 4 was diagnosed with talaromycosis marneffei involving the gastrointestinal tract. A colonoscopy revealed colon erosion, hyperemia, edema, and multiple intestinal mucosal ulcers (Figure 2A). Mucosal biopsy of transverse colon and descending colon tissue was performed (Figure 2B). After a period of antifungal treatment at our hospital, a repeat colonoscopy showed multiple ulcers in the colon in the healing stage, which indicated that the patient's condition had improved.\\nColonoscopy and histopathological examination. (A) The mucosa had apparent hyperemia and edema, as well as scattered erosion and ulceration, according to a colonoscopy. (B) Colon mucosa was stained with hexamine silver staining, showing several yeast‐like organisms with septate forms (indicated by the blue arrow).\",\n \"Six cases were diagnosed with talaromycosis marneffei by culture of bronchoalveolar lavage fluid and sputum specimens, blood microscopy, bronchoalveolar lavage fluid mNGS, intestinal lesion mNGS, and blood mNGS.\\n[SUBTITLE] Culture and microscopy observation [SUBSECTION] The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\nThe specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\\n[SUBTITLE] Metagenomic next‐generation sequencing [SUBSECTION] Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\\nAmong six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\",\n \"The specimens were inoculated into Sabouraud agar medium and incubated at 25°C and 37°C. The fungus appeared as a mold with red wine pigment at 25°C (Figure 3A), and curved conidial chains were evident under the microscope. The cultures were stained with lactophenol cotton blue, which revealed smooth transparent branched separated mycelia typical of broom‐like branches, with multiple scattered nonparallel peduncle bases with 2–10 vial peduncles on them and narrowed apically with scattered single chains of conidia (Figure 3B). It grew as yeast without pigment yield at 37°C, and fungal cells under the microscope presented as round or oval‐shaped yeast‐like cells with a diameter of 3 μm (Figure 2B). The fungus was identified by its characteristics. Peripheral blood smear was selected for direct microscopic examination after Wright–Giemsa staining. Round, oval, or sausage‐shaped mounded or scattered organisms could be seen in neutrophils and macrophages (Figure 4).\\nCulture and microscopy observation of Talaromyces marneffei. (A) Morphology and wine pigment of T. marneffei colonies from alveolar lavage fluid culture on Sabouraud agar medium for 7 days at 25°C. (B) Typical broom‐shaped, branching septate hyphae of T. marneffei at 25°C with lactophenol cotton blue staining (×400).\\nBlood smear examination. Blood smears were stained with Wright‐Giemsa staining, which showed small round‐to‐ovoid yeast cells within the neutrophil cytoplasm. Yeast cells were unevenly stained and showed strong staining at one end (indicated by the blue arrow) (×1000).\",\n \"Among six patients, only two patients were diagnosed with talaromycosis marneffei by mNGS. Due to exacerbation of the condition and multiple negative sample cultures, clinicians collected related samples for mNGS. mNGS of the BALF of case 3 revealed 38 T. marneffei reads, and mNGS of the intestinal lesion of case 4 revealed 27,935 T. marneffei reads. In addition, the presence of T. marneffei was also detected in their blood by mNGS (Table 5).\\nMicroorganism detected by mNGS of two patients\\nAbbreviations: BALF, bronchoalveolar lavage fluid; mNGS, metagenomic next‐generation sequencing.\",\n \"Amphotericin B, itraconazole, and voriconazole are routinely utilized for talaromycosis marneffei in clinical treatment. In terms of clinical response and fungicidal action, amphotericin B was superior to itraconazole in the initial treatment.\\n7\\n Voriconazole was reported to be effective for disseminated T. marneffei infection.\\n8\\n Thus, most of the cases in this study were treated with intravenous amphotericin B at a dose of 0.6 mg/kg/day for 2 weeks, followed by intravenous voriconazole at a dose of 200 mg/day for maintenance therapy. One mild case was treated with intravenous voriconazole at a dose of 200 mg q12 h for the first 24 h, followed by oral therapy at 200 mg twice daily. To reduce the incidence of adverse effects, liver function, renal function, and body electrolytes were monitored closely and regularly during treatment.\\nDuring the treatment period, adverse drug reactions to amphotericin B were observed. Two cases showed elevated creatinine levels. These patients were treated with lower doses of amphotericin B, or the drug was discontinued and replaced with voriconazole therapy. Only one patient did not receive antifungal medication and died within 24 h of admission due to deterioration of the disease. During the 6‐month follow‐up, most of the cases in this study showed improvement after treatment and had a good prognosis. One case is still under treatment and follow‐up.\",\n \"\\nTalaromyces marneffei is a species of penicillium isolated from the liver of bamboo rats that died unnaturally in Vietnam in 1956. In 1959, the fungus was named Penicillium marneffei in honor of Hubert Marneffe, the director of the Pasteur Institute in Indochina.\\n9\\n In 2011, based on phylogenetic and phenotypic analysis, the name was changed to T. marneffei.\\n10\\n Bamboo rats are currently considered natural hosts or carriers of this pathogen.\\n11\\n Regarding the patients' epidemiological history, it was found that no patients had a history of residence in endemic areas or exposure to bamboo rats. The original route of infection is currently unknown, but it could be through skin lesions or conidia inhalation. It has been documented that T. marneffei enters the blood system by invading the reticuloendothelial system and then spreads to other organs through the lymphatic system and blood circulation.\\n12\\n, \\n13\\n\\nT. marneffei mainly invades organs and sites related to the monocyte–macrophage reticuloendothelial system. T. marneffei infections can be divided into disseminated and focal types; focal type infections are often confined to the site of invasion, and the clinical manifestations are dominated by the symptoms of the original disease. The disseminated type often involves multiple tissues and organs. Common clinical manifestations of disseminated T. marneffei include fever, respiratory signs, anemia, weight loss, skin lesions, lymphadenopathy, and hepatosplenomegaly.\\n3\\n In this study, five patients exhibited disseminated T. marneffei infections. These patients presented with fever, cough, lymph node enlargement, anemia, and gastrointestinal symptoms, all of which were nonspecific and led to a misdiagnosis of tuberculosis or other fungal infections. Skin lesions are a clinical feature of disseminated T. marneffei, and characteristic lesions are papules with central necrosis.\\n9\\n Two patients in our cohort presented with atypical skin lesions, such as subcutaneous nodules or abscesses. In three cases, the lungs were affected by fungal invasion first. In addition, the possibility of gastrointestinal infection was not excluded. Although lymphoid tissue is extensive throughout the digestive tract, talaromycosis marneffei infection of the intestine is uncommon.\\n13\\n The atypical and nonspecific clinical presentation of fungal infections in T. marneffei often adds to the difficulty of early diagnosis, resulting in diagnostic delay and increased mortality.\\nT lymphocyte cells are critical for controlling T. marneffei infection.\\n14\\n Some experiments have indicated that CD4+ T cells mediate significant pulmonary inflammatory infiltration in mice by recruiting macrophages and granulocytes. CD4 deficiency may lead to the loss of Th2 anti‐inflammatory cytokines (IL‐4, IL‐5, and IL‐10).\\n15\\n In the innate immune system, activated macrophages can activate some proinflammatory cytokines and chemokines.\\n16\\n Myeloperoxidase‐dependent neutrophils also have fungicidal activity through the actions of reactive oxygen species.\\n17\\n In HIV‐positive individuals, T. marneffei‐activated monocyte‐derived dendritic cells (MDDCs) may further accelerate immunosuppression by promoting HIV‐1 transfection of CD4+ T cells.\\n18\\n Overall, the inflammatory and regulatory responses to a fungal infection are significantly modulated by the innate and acquired immune systems. Lymphocyte subpopulation analysis should be performed in clinical work, especially in immunosuppressed or HIV‐positive patients.\\nThe pathogenesis of T. marneffei involves the ability of the fungus to avoid being killed by macrophages and replicate inside them.\\n19\\n\\nT. marneffei can produce melanin or melanin‐like compounds to enhance resistance to phagocytosis by macrophages.\\n20\\n After phagocytosis by macrophages, T. marneffei (as an intracellular yeast form) induces the production of superoxide dismutase and catalase‐peroxidase in response to oxidative stress.\\n19\\n Other factors may also be associated with the pathogenicity of T. marneffei, including heat shock proteins (Hsp70 and Hsp30),\\n19\\n secreted galactomannan protein Mp1p\\n21\\n and the expression of isocitrate lyase.\\n22\\n Subsequently, conidial replication and establishment of infection mediate disseminated infections.\\nThe incidence of infection in HIV‐negative patients is increasing. For example, in 2012, Yongxuan Hu reviewed 668 cases of talaromycosis marneffei in mainland China from January 1984 to December 2009, and there were 82 (12.3%) HIV‐negative infected patients.\\n3\\n In 2015, Ye Qiu retrospectively analyzed 109 patients admitted to the First Affiliated Hospital of Guangxi Medical University for talaromycosis marneffei from January 1, 2003, to August 1, 2014, 39.45% of whom were HIV‐negative.\\n23\\n In HIV‐negative individuals, infection factors are most likely attributed to underlying conditions, including immunodeficiency due to anti‐IFN‐γ autoantibodies, immunosuppressive therapy, malignancy, diabetes mellitus, and organ transplantation.\\n24\\n Anti‐IFN‐γ autoantibodies can inhibit STAT1 phosphorylation and IL‐12 production, increasing the risk of fungal infection.\\n25\\n Other predisposing factors may be related to lymphocyte depletion and impairment of lymphocyte proliferation. Interestingly, one of our cases was an immunocompetent patient, and the infection factor was most likely attributed to the patient's diabetes (Table 2).\\nIn terms of laboratory tests, all patients had elevated serum CRP levels and erythrocyte sedimentation rates (Table 1). Some patients had liver damage, characterized by elevated aspartate aminotransferase (AST), especially severe patients. HIV‐positive patients were more likely to have leukopenia, low platelet counts, elevation of alanine transaminase, and positive blood cultures. In addition, the chest imaging manifestations of patients with T. marneffei infection were diverse, including multiple patches or large consolidations, ground‐glass opacities, pleural effusion, and hilar and mediastinal lymph node enlargement (Table 1). Infectious diseases such as tuberculosis, Aspergillus pneumonia, lymphoma, lung cancer, and others should be ruled out.\\n26\\n\\n\\nThe traditional methods for diagnosis of talaromycosis marneffei include culture of the pathogen and histopathological examination.\\n1\\n However, the culture positivity rate was not high in patients with early infection. These sample culture results were presented as repeated negative, including a blood culture of case 3 and a stool culture of case 4. Among the six cases, two were diagnosed with talaromycosis marneffei by mNGS. One case was diagnosed with talaromycosis marneffei by examination of a peripheral blood smear (Figure 4). Yeast‐like organisms may be seen on the blood smear, especially in patients with heavy fungemia.\\n27\\n Blood culture was positive in only two patients in our study (Table 2). Whereas it usually takes 3–10 days for a patient's clinical specimen to be cultured for T. marneffei, the time from admission to diagnosis was long for most patients in our study. During this period, some patients were misdiagnosed with bacterial pneumonia or other diseases. Moreover, they received improper treatment. When none of their conditions improved significantly, physicians began to suspect the initial diagnosis and considered mNGS. Other techniques, such as novel Mp1p enzyme immunoassay and the beta‐D‐glucan test, have a certain role in the auxiliary diagnosis of talaromycosis marneffei.\\n28\\n, \\n29\\n However, these approaches have the drawbacks of extensive processes and limited sensitivity. Several real‐time PCR techniques to identify T. marneffei have been developed in response to the demand for faster and more sensitive diagnostics, although these are limited to primer and probe design.\\n30\\n, \\n31\\n Under such circumstances, the implementation of NGS in the clinical field might have some undeniable advantages.\\nmNGS, a rapid and effective diagnostic approach, can obtain information on microbial species and abundance in samples by high‐throughput sequencing of nucleic acids and bioinformatics analysis.\\n32\\n It not only comprehensively identifies pathogens but also offers a significantly reduced time requirement for pathogen identification. In this study, clinicians chose mNGS in the following situations: (1) difficulty identifying fungal pathogens despite repeated cultures and (2) a need for early and precise diagnosis in patients with severe pathogenic infections. Multiple studies have reported the use of mNGS in clinical fungal diagnosis, which also reveals the potential merits of this technique in rapid etiological diagnosis.\\n32\\n, \\n33\\n, \\n34\\n, \\n35\\n, \\n36\\n However, there are some shortcomings in the widespread application of mNGS, such as high cost and a lack of recognized standards in this area.\\nTalaromycosis marneffei is an uncommon disease that causes high mortality rates. A previous study showed that HIV infection, CD4/CD8 < 0.5, a reduced percentage of CD4+ T cells and late diagnosis were potential risk factors for poor prognosis.\\n37\\n Late diagnosis is the leading cause of death in T. marneffei‐infected individuals.\\n38\\n It is clear that early and rapid diagnosis is particularly important. Therefore, clinicians should consider talaromycosis marneffei if a patient has suffered recurrent fever; multiple patches, nodules, or masses in the lungs; gastrointestinal symptoms; and characteristic skin lesions, especially if anti‐inflammatory treatment is ineffective. Pathogen cultivation and other relevant laboratory tests are essential. In addition, we recommend the early use of mNGS to reduce mortality and prolong survival time. There are some limitations of this study. First, it is a retrospective cohort study, and the limited number of cases makes it difficult to explore the differences in clinical manifestations between HIV‐positive and HIV‐negative populations. Second, there is a lack of comprehensive analysis of the infection mechanism of T. marneffei in nonendemic areas.\\nIn conclusion, we analyzed six cases of talaromycosis marneffei in nonendemic areas in the hope of promoting awareness and familiarity with this rare disease among more clinicians in nonendemic areas and improving the prognosis of patients.\",\n \"All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Lei Peng and Xing‐bei Weng. The first draft of the article was written by Lei Peng, and all authors commented on previous versions of the article. All authors read and approved the final article.\",\n \"This work was supported by the Ningbo Natural Science Foundation (2019A610381) and the Ningbo Public Welfare Foundation (2019C50087).\",\n \"The authors declare no competing interests.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,"methods",null,null,null,null,null,null,"results",null,null,null,null,null,null,null,"discussion",null,null,"COI-statement"],"string":"[\n null,\n \"methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n \"COI-statement\"\n]"},"keywords":{"kind":"list like","value":["disseminated Talaromyces marneffei infection","laboratory test","metagenomic next‐generation sequencing","\nTalaromyces marneffei\n","talaromycosis marneffei"],"string":"[\n \"disseminated Talaromyces marneffei infection\",\n \"laboratory test\",\n \"metagenomic next‐generation sequencing\",\n \"\\nTalaromyces marneffei\\n\",\n \"talaromycosis marneffei\"\n]"}}},{"rowIdx":386,"cells":{"title":{"kind":"string","value":"An International Consensus on Actions to Improve Lung Cancer Survival: A Modified Delphi Method Among Clinical Experts in the International Cancer Benchmarking Partnership."},"pmid":{"kind":"string","value":"36269109"},"background_abstract":{"kind":"string","value":"Research from the International Cancer Benchmarking Partnership (ICBP) demonstrates that international variation in lung cancer survival persists, particularly within early stage disease. There is a lack of international consensus on the critical contributing components to variation in lung cancer outcomes and the steps needed to optimise lung cancer services. These are needed to improve the quality of options for and equitable access to treatment, and ultimately improve survival."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Semi-structured interviews were conducted with 9 key informants from ICBP countries. An international clinical network representing 6 ICBP countries (Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland & Wales) was established to share local clinical insights and examples of best practice. Using a modified Delphi consensus model, network members suggested and rated recommendations to optimise the management of lung cancer. Calls to Action were developed via Delphi voting as the most crucial recommendations, with Good Practice Points included to support their implementation."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Five Calls to Action and thirteen Good Practice Points applicable to high income, comparable countries were developed and achieved 100% consensus. Calls to Action include (1) Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives; (2) Ensure diagnosis of lung cancer within 30 days of referral; (3) Develop Thoracic Centres of Excellence; (4) Undertake an international audit of lung cancer care; and (5) Recognise improvements in lung cancer care and outcomes as a priority in cancer policy."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The recommendations presented are the voice of an expert international lung cancer clinical network, and signpost key considerations for policymakers in countries within the ICBP but also in other comparable high-income countries. These define a roadmap to help align and focus efforts in improving outcomes and management of lung cancer patients globally."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Benchmarking","Lung Neoplasms","Consensus","Early Detection of Cancer","Delphi Technique"],"string":"[\n \"Humans\",\n \"Benchmarking\",\n \"Lung Neoplasms\",\n \"Consensus\",\n \"Early Detection of Cancer\",\n \"Delphi Technique\"\n]"},"pmcid":{"kind":"string","value":"9596933"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Lung cancer accounts for nearly 1.8 million deaths annually – more than colorectal and breast cancers combined.1 Despite significant improvements in diagnosis and treatment, lung cancer is typically associated with low survival internationally.2 The International Cancer Benchmarking Partnership (ICBP) is a global collaboration of clinicians, policymakers, researchers, and cancer data experts. It seeks to benchmark and explain cancer survival, incidence, and mortality differences between high-income countries with comprehensive cancer registry coverage, similar budgetary spending on national health systems and universal access to healthcare. Recent ICBP research has demonstrated that international disparities in lung cancer survival persist, with 5-year survival ranging from 14.7% (UK) to 21.7% (Canada).2 Evidence shows that 3-year survival varies mostly in potentially curative lung cancer, by up to 20 percentage points for localised and regional disease (as per SEER summary staging)3 across the ICBP countries.4 Comparably, survival varied by just 3.4 percentage points in patients diagnosed with distant/metastatic disease across the same countries.4\nDue to the inherent complexity of health systems, it is unlikely that one single component of the cancer care continuum can ultimately define why survival variation exists, yet the ability to diagnose and stage lung cancer patients as early as possible will likely influence their treatment options.5 Various work has been undertaken to articulate patient, healthcare professional and disease related factors that influence timeliness of diagnosis and ultimately, patient outcomes.6 This understanding has helped build initiatives across the pathway to address the poorer outcomes in lung cancer, but to differing levels internationally. This starts with preventative initiatives for lung cancer, including support for smoking cessation and lung cancer awareness campaigns. Both areas are of significance, with tobacco being a key factor in the aetiology of lung cancer, and some awareness campaigns influencing stage shift.7,8 Variation in primary care practitioner readiness to refer patients with potential symptoms and in service provision of key diagnostic and staging tools for lung cancer, such as PET-CT, have been demonstrated across ICBP countries with unknown quantifiable impacts upon survival.9,10 This is reflected in the variation seen in lung cancer stage distribution between ICBP countries, but not to the same extent as differences in survival by stage.4 In the diagnostic phase, it is clear that international differences exist for lung cancer patients, yet the evidence for how this affects outcomes is variable and complex.11 Variation within the diagnosis and staging of early stage lung cancer patients is important; unwarranted delays enabling tumour growth may shift patients away from being in a potentially curable stage towards more advanced stages with lower survival.12,13\nCurative intent treatment options for lung cancer have increased and become significantly more sophisticated over recent years, with the development and adoption of more precise radiotherapy, and minimally invasive surgery.14 Less invasive options have opened up better opportunities for more complex patients or those with greater comorbidities. Guidance and uptake of these treatment options for early stage patients is likely to vary both within countries and internationally. This is particularly the case in patients deemed to be borderline candidates for curative therapy where there are more variables playing a role in decision-making for treatment.15 Variation in the use of curative intent therapies within early stage patients has been demonstrated in some countries.16 Disease stage, patient fitness/performance status, lung function, staffing skillset, socioeconomic status and comorbidities play key influential roles in the perceived operability of patients and subsequent treatment decisions.16\nThe organisation of each country’s healthcare system has been established to guide patients from presentation to the decision for treatment, with many informal and formal treatment pathways being established. However, gaps and biases within each system have been identified that undoubtedly affect the management of potentially curable patients.17 There is a current lack of internationally agreed recommendations for improving care of lung cancer patients, with the purpose of being interpreted at a policy level to align efforts to empower those in positions to instigate change.\nWithin this study, we have established an international lung cancer network to share local clinical insights, including locally defined best practices, mitigation of modifiable variables and opportunities for collaboration. We present a series of recommendations to improve lung cancer care, in order to improve survival, based on the consensus of our international clinical network. We hope that these recommendations serve to optimise lung cancer services so as to improve the quality of options for treatment, equitable access to treatment, and ultimately improve survival internationally."},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Initial Interview Results [SUBSECTION] These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\nThese are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\n[SUBTITLE] Recommendations [SUBSECTION] A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\nA total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\n[SUBTITLE] Calls to Action Summary [SUBSECTION] Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nCalls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Good Practice Points Summary [SUBSECTION] The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\nThe network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics."},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"The burden of lung cancer is significant and represents a global public health challenge. The Calls to Action and Good Practice Points presented here, reflecting the views of an international expert lung cancer clinical network, signpost the key considerations for policymakers within high-income comparable countries to effective change. They provide a roadmap to help align and focus efforts in improving outcomes and management of lung cancer patients. The Good Practice Points described are key to enable delivery of the Calls to Action, which should be considered the highest priority. Collaboration at local, regional, national, and international levels within the lung cancer community is imperative to empower policymakers to ensure policies are enacted that enhance lung cancer patient treatment and improve quality of life and patient outcomes."},"other_sections_titles":{"kind":"list like","value":["Network Formation","Topic Guide Development and Initial Interviews","Roundtable Discussions and Recommendation Development","Modified Delphi Method","Refinement of Recommendations","Initial Interview Results","Recommendations","Calls to Action Summary","Box 1A: Calls to Action","Good Practice Points Summary","Box 1B: Good Practice Points","Pre-Diagnosis","Diagnosis","Therapeutic","All Phases of Care","Implement Cost-Effective, Clinically Efficacious, and Equitable Lung Cancer Screening Initiatives","Ensure Diagnosis of Lung Cancer Within 30 days of Referral","Develop Thoracic Centres of Excellence","Undertake an International Audit of Lung Cancer Care","Recognise Improvements in Lung Cancer Care and Outcomes as a Priority in Cancer Policy","Limitations"],"string":"[\n \"Network Formation\",\n \"Topic Guide Development and Initial Interviews\",\n \"Roundtable Discussions and Recommendation Development\",\n \"Modified Delphi Method\",\n \"Refinement of Recommendations\",\n \"Initial Interview Results\",\n \"Recommendations\",\n \"Calls to Action Summary\",\n \"Box 1A: Calls to Action\",\n \"Good Practice Points Summary\",\n \"Box 1B: Good Practice Points\",\n \"Pre-Diagnosis\",\n \"Diagnosis\",\n \"Therapeutic\",\n \"All Phases of Care\",\n \"Implement Cost-Effective, Clinically Efficacious, and Equitable Lung Cancer Screening Initiatives\",\n \"Ensure Diagnosis of Lung Cancer Within 30 days of Referral\",\n \"Develop Thoracic Centres of Excellence\",\n \"Undertake an International Audit of Lung Cancer Care\",\n \"Recognise Improvements in Lung Cancer Care and Outcomes as a Priority in Cancer Policy\",\n \"Limitations\"\n]"},"other_sections_texts":{"kind":"list like","value":["Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.","We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.","Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.","Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\nStepwise modified Delphi Method.","‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.","These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.","A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).","Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy","Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy","The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.","• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.","• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.","• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","Due to variation seen in stage distribution between ICBP countries, it is important to consider where opportunities exist to optimise care and improve survival in the pre-treatment interval. Whilst the evidence is strong for a survival benefit of low dose CT screening in individuals with higher risk of lung cancer, implementation has been a complex issue for policymakers.22-24 Whilst low dose CT screening can support detection of early stage disease and increase the likelihood of curable treatment, several factors must be carefully considered to ensure clinical efficacy and cost effectiveness. Integration into existing infrastructure, target populations, recruitment of population at risk, reduction of harm, resource and cost, and engagement with the public and healthcare professionals need to be defined and implemented. Screening programmes need to be designed to reach a traditionally ‘hard to reach’ population, requiring their design to have greater engagement of those affected. There is no “one size fits all” approach, yet it is the collective opinion of this international clinical network that understanding the impact that screening could have for outcomes and considering all these factors within local contexts is a crucial activity for policymakers. Currently, no ICBP country has fully implemented a national lung cancer screening programme, though a number of countries are at different stages of exploration and implementation. Successful models of assessment and pilot trials have been underway internationally for several years and can serve as a baseline to inform the best implementation plan for screening in each jurisdiction.25-29","Various temporal targets for key time points within the cancer patient journey exist for different countries, but there has previously been a lack of international consensus on a universal ‘best practice’ target. Unwarranted delays in diagnosis can rapidly shift lung cancer patients from being potentially curative to being difficult to treat, highlighting the importance of timely diagnosis to improve the possibility of treatment with curative intent.12 Additionally, differences in the diagnostic interval for lung cancer has been demonstrated across ICBP countries, supporting the need for greater alignment in international efforts to improve time to diagnosis and treatment parameters.30 Different countries take a more aggressive stance in their time targets with compulsory reporting, and the accountability for these (e.g. Denmark), whilst in others this can be less regimented.31 The network formed in this study have agreed that <30 days from point of referral for possible lung cancer diagnosis should be a minimum requirement to allow best opportunity for timely initiation of treatment. A crucial consideration within this target is the need to ensure timely access to diagnostic and staging tools, with strong coordination between diagnostic and treatment services – the need to improve this from a policy and practice perspective will differ, depending on jurisdictional context.","Developing Thoracic Centres of Excellence and formally affiliated networks can help provide advanced and comprehensive patient-focused approaches to lung cancer treatment. Implementation of this aspiration in different regions may differ but essential components of Centres of Excellence for lung cancer have been researched and agreed by this international network. Expertise from multiple disciplines dedicated to lung cancer management (including thoracic surgery, oncology, anaesthesia, respiratory medicine, pathology, radiology, pulmonology rehabilitation) should be available within the Centre and its network. This enables seamless and multi-disciplinary care for patients. Critical to this is ensuring that robustly trained healthcare professionals and specialist experts relevant to these disciplines are deployed, with investment in their training and development to support continued improvement. Broader formal networks are required to provide contemporary access for patients, alongside streamlined processes across prevention, diagnosis, treatment, and survivorship care. It is important that Centres provide comprehensive care to all patients, including those first referred to local hospitals. There is some evidence in the UK that first referral to a local hospital is associated with fewer patients being offered curative treatment.32 The explanation for this is not clear but may relate to distances that patients need to travel.32 Centres should support availability of clinical trials to help deliver innovative research advances to treatments and to enhance patient outcomes. Ultimately, Thoracic Centres of Excellence and their networks should be viewed as the backbone of timely integrated care for lung cancer patients.18 Development of such Centres has been underway across ICBP countries, but relies heavily on available resource, funding, capacity, human resources, and coordination. When considering service improvement and expansion, Centres of Excellence should be in the forefront of policy plans and considerations.","High-quality data capture and timely feedback is crucial to improving care.33 The importance of this intelligence at local, regional, and national levels needs to be recognised across all countries. Embedding data capture and feedback loops into existing healthcare structures and processes in lung cancer care should be prioritised for self-evaluation and to drive service improvements. Whilst some work has been undertaken to better understand variation in delivery of lung cancer care between comparable high-income countries (as described in the Background), a truly comprehensive understanding of the key drivers of better outcomes in certain countries requires a detailed international audit. In order to deliver this vital intelligence, a number of considerations re required. A universal minimum dataset with key metrics (wait time, stage at diagnosis, risk adjusted resection and radiation rates, medical oncology treatments, significant adverse event rates) is required, alongside strong collaborative relationships between participating countries.34 Ensuring the minimum data metrics are collected as part of routine practice should be the first step to enable a successful audit within jurisdictions, in order to inform the delivery of an international audit.35 Few national audits have been undertaken for lung cancer care, but if more countries replicate this, the roadmap to commissioning an international audit will be clearer.36-38 This will allow clear identification of international best practice to facilitate sharing of optimal care and lessons learnt.","Continuous improvement in lung cancer care and outcomes should be recognised as an expectation by policymakers, requiring ongoing investment. The success of cancer site-specific policies in high-income countries, for example breast cancer, is reflected in enhanced policy focus, driving service development (screening programmes, diagnostic initiatives and treatment options) and underpinning improved survival internationally.39 However, even in higher performing countries, lung cancer survival is still low.2,4 Significant inequities in lung cancer risk, access to care and treatment outcomes have historically existed, driven largely by socioeconomic inequalities, and thus prohibiting improvements in lung cancer survival seen for other cancer sites.40 Recognising and addressing these inequities and inequalities in care collectively needs to be prioritised within cancer policy internationally, facilitating service improvement to optimise management and care of lung cancer patients. In order to drive this improvement, jurisdictions must consider the effectiveness of their leadership (within both clinical and political arenas) to instigate change, prioritise connection to and collaboration with appropriate stakeholders to ensure lung cancer is promoted on political agendas, and have the means to lobby for and disseminate improved lung cancer policy robustly.41 By calling for lung cancer care to be a priority internationally, we aim to encourage policymakers to prioritise the most pressing initiatives required within their respective countries to have the greatest impact upon lung cancer outcomes.","It is acknowledged by the network and the authors that these recommendations were developed in the context of the ICBP country membership and may not fully represent the diversity of healthcare settings globally. We are aware the interviews undertaken to help inform the development of the recommendations were of a limited number, however we targeted these at stakeholders who would have a breadth of knowledge. Data saturation would therefore be unlikely to occur but we hope this format and level of questioning can now be replicated and built upon in other healthcare systems and countries to further develop our understanding of key factors in improving lung cancer care internationally. Both the network and authors hope for the recommendations developed to be applicable to other comparable high-income countries, who can interpret these in the context of their healthcare system structure and funding, cultures, socioeconomic situation, and other population-based needs. Additionally, the importance of rigorously evaluating new initiatives, policies, and the evidence base to support their implementation within local settings should not be underestimated – it is the aim of the network and the authors to bring these recommendations to the forefront of policy attention, in order for local assessment of the requirements needed to evaluate and implement."],"string":"[\n \"Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\",\n \"We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\",\n \"Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\",\n \"Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\\nStepwise modified Delphi Method.\",\n \"‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.\",\n \"These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\",\n \"A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\",\n \"Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\",\n \"Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\",\n \"The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\",\n \"• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\",\n \"• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\",\n \"• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"Due to variation seen in stage distribution between ICBP countries, it is important to consider where opportunities exist to optimise care and improve survival in the pre-treatment interval. Whilst the evidence is strong for a survival benefit of low dose CT screening in individuals with higher risk of lung cancer, implementation has been a complex issue for policymakers.22-24 Whilst low dose CT screening can support detection of early stage disease and increase the likelihood of curable treatment, several factors must be carefully considered to ensure clinical efficacy and cost effectiveness. Integration into existing infrastructure, target populations, recruitment of population at risk, reduction of harm, resource and cost, and engagement with the public and healthcare professionals need to be defined and implemented. Screening programmes need to be designed to reach a traditionally ‘hard to reach’ population, requiring their design to have greater engagement of those affected. There is no “one size fits all” approach, yet it is the collective opinion of this international clinical network that understanding the impact that screening could have for outcomes and considering all these factors within local contexts is a crucial activity for policymakers. Currently, no ICBP country has fully implemented a national lung cancer screening programme, though a number of countries are at different stages of exploration and implementation. Successful models of assessment and pilot trials have been underway internationally for several years and can serve as a baseline to inform the best implementation plan for screening in each jurisdiction.25-29\",\n \"Various temporal targets for key time points within the cancer patient journey exist for different countries, but there has previously been a lack of international consensus on a universal ‘best practice’ target. Unwarranted delays in diagnosis can rapidly shift lung cancer patients from being potentially curative to being difficult to treat, highlighting the importance of timely diagnosis to improve the possibility of treatment with curative intent.12 Additionally, differences in the diagnostic interval for lung cancer has been demonstrated across ICBP countries, supporting the need for greater alignment in international efforts to improve time to diagnosis and treatment parameters.30 Different countries take a more aggressive stance in their time targets with compulsory reporting, and the accountability for these (e.g. Denmark), whilst in others this can be less regimented.31 The network formed in this study have agreed that <30 days from point of referral for possible lung cancer diagnosis should be a minimum requirement to allow best opportunity for timely initiation of treatment. A crucial consideration within this target is the need to ensure timely access to diagnostic and staging tools, with strong coordination between diagnostic and treatment services – the need to improve this from a policy and practice perspective will differ, depending on jurisdictional context.\",\n \"Developing Thoracic Centres of Excellence and formally affiliated networks can help provide advanced and comprehensive patient-focused approaches to lung cancer treatment. Implementation of this aspiration in different regions may differ but essential components of Centres of Excellence for lung cancer have been researched and agreed by this international network. Expertise from multiple disciplines dedicated to lung cancer management (including thoracic surgery, oncology, anaesthesia, respiratory medicine, pathology, radiology, pulmonology rehabilitation) should be available within the Centre and its network. This enables seamless and multi-disciplinary care for patients. Critical to this is ensuring that robustly trained healthcare professionals and specialist experts relevant to these disciplines are deployed, with investment in their training and development to support continued improvement. Broader formal networks are required to provide contemporary access for patients, alongside streamlined processes across prevention, diagnosis, treatment, and survivorship care. It is important that Centres provide comprehensive care to all patients, including those first referred to local hospitals. There is some evidence in the UK that first referral to a local hospital is associated with fewer patients being offered curative treatment.32 The explanation for this is not clear but may relate to distances that patients need to travel.32 Centres should support availability of clinical trials to help deliver innovative research advances to treatments and to enhance patient outcomes. Ultimately, Thoracic Centres of Excellence and their networks should be viewed as the backbone of timely integrated care for lung cancer patients.18 Development of such Centres has been underway across ICBP countries, but relies heavily on available resource, funding, capacity, human resources, and coordination. When considering service improvement and expansion, Centres of Excellence should be in the forefront of policy plans and considerations.\",\n \"High-quality data capture and timely feedback is crucial to improving care.33 The importance of this intelligence at local, regional, and national levels needs to be recognised across all countries. Embedding data capture and feedback loops into existing healthcare structures and processes in lung cancer care should be prioritised for self-evaluation and to drive service improvements. Whilst some work has been undertaken to better understand variation in delivery of lung cancer care between comparable high-income countries (as described in the Background), a truly comprehensive understanding of the key drivers of better outcomes in certain countries requires a detailed international audit. In order to deliver this vital intelligence, a number of considerations re required. A universal minimum dataset with key metrics (wait time, stage at diagnosis, risk adjusted resection and radiation rates, medical oncology treatments, significant adverse event rates) is required, alongside strong collaborative relationships between participating countries.34 Ensuring the minimum data metrics are collected as part of routine practice should be the first step to enable a successful audit within jurisdictions, in order to inform the delivery of an international audit.35 Few national audits have been undertaken for lung cancer care, but if more countries replicate this, the roadmap to commissioning an international audit will be clearer.36-38 This will allow clear identification of international best practice to facilitate sharing of optimal care and lessons learnt.\",\n \"Continuous improvement in lung cancer care and outcomes should be recognised as an expectation by policymakers, requiring ongoing investment. The success of cancer site-specific policies in high-income countries, for example breast cancer, is reflected in enhanced policy focus, driving service development (screening programmes, diagnostic initiatives and treatment options) and underpinning improved survival internationally.39 However, even in higher performing countries, lung cancer survival is still low.2,4 Significant inequities in lung cancer risk, access to care and treatment outcomes have historically existed, driven largely by socioeconomic inequalities, and thus prohibiting improvements in lung cancer survival seen for other cancer sites.40 Recognising and addressing these inequities and inequalities in care collectively needs to be prioritised within cancer policy internationally, facilitating service improvement to optimise management and care of lung cancer patients. In order to drive this improvement, jurisdictions must consider the effectiveness of their leadership (within both clinical and political arenas) to instigate change, prioritise connection to and collaboration with appropriate stakeholders to ensure lung cancer is promoted on political agendas, and have the means to lobby for and disseminate improved lung cancer policy robustly.41 By calling for lung cancer care to be a priority internationally, we aim to encourage policymakers to prioritise the most pressing initiatives required within their respective countries to have the greatest impact upon lung cancer outcomes.\",\n \"It is acknowledged by the network and the authors that these recommendations were developed in the context of the ICBP country membership and may not fully represent the diversity of healthcare settings globally. We are aware the interviews undertaken to help inform the development of the recommendations were of a limited number, however we targeted these at stakeholders who would have a breadth of knowledge. Data saturation would therefore be unlikely to occur but we hope this format and level of questioning can now be replicated and built upon in other healthcare systems and countries to further develop our understanding of key factors in improving lung cancer care internationally. Both the network and authors hope for the recommendations developed to be applicable to other comparable high-income countries, who can interpret these in the context of their healthcare system structure and funding, cultures, socioeconomic situation, and other population-based needs. Additionally, the importance of rigorously evaluating new initiatives, policies, and the evidence base to support their implementation within local settings should not be underestimated – it is the aim of the network and the authors to bring these recommendations to the forefront of policy attention, in order for local assessment of the requirements needed to evaluate and implement.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Materials and Methods","Network Formation","Topic Guide Development and Initial Interviews","Roundtable Discussions and Recommendation Development","Modified Delphi Method","Refinement of Recommendations","Results","Initial Interview Results","Recommendations","Calls to Action Summary","Box 1A: Calls to Action","Good Practice Points Summary","Box 1B: Good Practice Points","Pre-Diagnosis","Diagnosis","Therapeutic","All Phases of Care","Discussion","Implement Cost-Effective, Clinically Efficacious, and Equitable Lung Cancer Screening Initiatives","Ensure Diagnosis of Lung Cancer Within 30 days of Referral","Develop Thoracic Centres of Excellence","Undertake an International Audit of Lung Cancer Care","Recognise Improvements in Lung Cancer Care and Outcomes as a Priority in Cancer Policy","Limitations","Conclusion","Supplemental Material"],"string":"[\n \"Introduction\",\n \"Materials and Methods\",\n \"Network Formation\",\n \"Topic Guide Development and Initial Interviews\",\n \"Roundtable Discussions and Recommendation Development\",\n \"Modified Delphi Method\",\n \"Refinement of Recommendations\",\n \"Results\",\n \"Initial Interview Results\",\n \"Recommendations\",\n \"Calls to Action Summary\",\n \"Box 1A: Calls to Action\",\n \"Good Practice Points Summary\",\n \"Box 1B: Good Practice Points\",\n \"Pre-Diagnosis\",\n \"Diagnosis\",\n \"Therapeutic\",\n \"All Phases of Care\",\n \"Discussion\",\n \"Implement Cost-Effective, Clinically Efficacious, and Equitable Lung Cancer Screening Initiatives\",\n \"Ensure Diagnosis of Lung Cancer Within 30 days of Referral\",\n \"Develop Thoracic Centres of Excellence\",\n \"Undertake an International Audit of Lung Cancer Care\",\n \"Recognise Improvements in Lung Cancer Care and Outcomes as a Priority in Cancer Policy\",\n \"Limitations\",\n \"Conclusion\",\n \"Supplemental Material\"\n]"},"all_sections_texts":{"kind":"list like","value":["Lung cancer accounts for nearly 1.8 million deaths annually – more than colorectal and breast cancers combined.1 Despite significant improvements in diagnosis and treatment, lung cancer is typically associated with low survival internationally.2 The International Cancer Benchmarking Partnership (ICBP) is a global collaboration of clinicians, policymakers, researchers, and cancer data experts. It seeks to benchmark and explain cancer survival, incidence, and mortality differences between high-income countries with comprehensive cancer registry coverage, similar budgetary spending on national health systems and universal access to healthcare. Recent ICBP research has demonstrated that international disparities in lung cancer survival persist, with 5-year survival ranging from 14.7% (UK) to 21.7% (Canada).2 Evidence shows that 3-year survival varies mostly in potentially curative lung cancer, by up to 20 percentage points for localised and regional disease (as per SEER summary staging)3 across the ICBP countries.4 Comparably, survival varied by just 3.4 percentage points in patients diagnosed with distant/metastatic disease across the same countries.4\nDue to the inherent complexity of health systems, it is unlikely that one single component of the cancer care continuum can ultimately define why survival variation exists, yet the ability to diagnose and stage lung cancer patients as early as possible will likely influence their treatment options.5 Various work has been undertaken to articulate patient, healthcare professional and disease related factors that influence timeliness of diagnosis and ultimately, patient outcomes.6 This understanding has helped build initiatives across the pathway to address the poorer outcomes in lung cancer, but to differing levels internationally. This starts with preventative initiatives for lung cancer, including support for smoking cessation and lung cancer awareness campaigns. Both areas are of significance, with tobacco being a key factor in the aetiology of lung cancer, and some awareness campaigns influencing stage shift.7,8 Variation in primary care practitioner readiness to refer patients with potential symptoms and in service provision of key diagnostic and staging tools for lung cancer, such as PET-CT, have been demonstrated across ICBP countries with unknown quantifiable impacts upon survival.9,10 This is reflected in the variation seen in lung cancer stage distribution between ICBP countries, but not to the same extent as differences in survival by stage.4 In the diagnostic phase, it is clear that international differences exist for lung cancer patients, yet the evidence for how this affects outcomes is variable and complex.11 Variation within the diagnosis and staging of early stage lung cancer patients is important; unwarranted delays enabling tumour growth may shift patients away from being in a potentially curable stage towards more advanced stages with lower survival.12,13\nCurative intent treatment options for lung cancer have increased and become significantly more sophisticated over recent years, with the development and adoption of more precise radiotherapy, and minimally invasive surgery.14 Less invasive options have opened up better opportunities for more complex patients or those with greater comorbidities. Guidance and uptake of these treatment options for early stage patients is likely to vary both within countries and internationally. This is particularly the case in patients deemed to be borderline candidates for curative therapy where there are more variables playing a role in decision-making for treatment.15 Variation in the use of curative intent therapies within early stage patients has been demonstrated in some countries.16 Disease stage, patient fitness/performance status, lung function, staffing skillset, socioeconomic status and comorbidities play key influential roles in the perceived operability of patients and subsequent treatment decisions.16\nThe organisation of each country’s healthcare system has been established to guide patients from presentation to the decision for treatment, with many informal and formal treatment pathways being established. However, gaps and biases within each system have been identified that undoubtedly affect the management of potentially curable patients.17 There is a current lack of internationally agreed recommendations for improving care of lung cancer patients, with the purpose of being interpreted at a policy level to align efforts to empower those in positions to instigate change.\nWithin this study, we have established an international lung cancer network to share local clinical insights, including locally defined best practices, mitigation of modifiable variables and opportunities for collaboration. We present a series of recommendations to improve lung cancer care, in order to improve survival, based on the consensus of our international clinical network. We hope that these recommendations serve to optimise lung cancer services so as to improve the quality of options for treatment, equitable access to treatment, and ultimately improve survival internationally.","[SUBTITLE] Network Formation [SUBSECTION] Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\nNine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\n[SUBTITLE] Topic Guide Development and Initial Interviews [SUBSECTION] We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\nWe initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\n[SUBTITLE] Roundtable Discussions and Recommendation Development [SUBSECTION] Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\nFour roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\n[SUBTITLE] Modified Delphi Method [SUBSECTION] Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\nStepwise modified Delphi Method.\nNetwork members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\nStepwise modified Delphi Method.\n[SUBTITLE] Refinement of Recommendations [SUBSECTION] ‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.\n‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.","Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.","We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.","Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.","Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\nStepwise modified Delphi Method.","‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.","[SUBTITLE] Initial Interview Results [SUBSECTION] These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\nThese are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\n[SUBTITLE] Recommendations [SUBSECTION] A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\nA total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\n[SUBTITLE] Calls to Action Summary [SUBSECTION] Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nCalls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Good Practice Points Summary [SUBSECTION] The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\nThe network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.","A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).","Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy","Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\nEnsure diagnosis of lung cancer within 30 days of referral\nDevelop Thoracic Centres of Excellence\nUndertake an international audit of lung cancer care\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy","The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.","• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\n• Protocols for incorporation into workplans, coordination and chairing of meetings\n• Appropriate attendance from core stakeholders\n• Means to record and report treatment decisions locally and centrally\n• Method to communicate treatment decisions to referring practitioners\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.","• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.","• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.","Rapid progress in areas of healthcare need can be achieved when international collaborative efforts, research and funding are aligned to a common goal, most recently illustrated by the development of the COVID vaccine.20 The international cancer community recognises the need to significantly improve lung cancer survival.21 The Calls to Action and Good Practice Points generated in this study provide a baseline for reflection, appropriate benchmarking, and identification of opportunities for enhancing lung cancer care. These can be interpreted from local, regional, national, and international perspectives. Many advances have been made in lung cancer prevention, screening, diagnosis, and treatment worldwide – from lower smoking rates, utilisation of CT scans for earlier detection, greater access to innovative medicines, improved operative techniques and more advanced radiation therapy. However, the potential improvements in survival have not yet been achieved. The international network formed, and recommendations developed within this study, are advocating for increased and better alignment of efforts within, and across, countries in order to improve lung cancer care and survival. We outline the importance and considerations to be had when addressing the Calls to Action below.","Due to variation seen in stage distribution between ICBP countries, it is important to consider where opportunities exist to optimise care and improve survival in the pre-treatment interval. Whilst the evidence is strong for a survival benefit of low dose CT screening in individuals with higher risk of lung cancer, implementation has been a complex issue for policymakers.22-24 Whilst low dose CT screening can support detection of early stage disease and increase the likelihood of curable treatment, several factors must be carefully considered to ensure clinical efficacy and cost effectiveness. Integration into existing infrastructure, target populations, recruitment of population at risk, reduction of harm, resource and cost, and engagement with the public and healthcare professionals need to be defined and implemented. Screening programmes need to be designed to reach a traditionally ‘hard to reach’ population, requiring their design to have greater engagement of those affected. There is no “one size fits all” approach, yet it is the collective opinion of this international clinical network that understanding the impact that screening could have for outcomes and considering all these factors within local contexts is a crucial activity for policymakers. Currently, no ICBP country has fully implemented a national lung cancer screening programme, though a number of countries are at different stages of exploration and implementation. Successful models of assessment and pilot trials have been underway internationally for several years and can serve as a baseline to inform the best implementation plan for screening in each jurisdiction.25-29","Various temporal targets for key time points within the cancer patient journey exist for different countries, but there has previously been a lack of international consensus on a universal ‘best practice’ target. Unwarranted delays in diagnosis can rapidly shift lung cancer patients from being potentially curative to being difficult to treat, highlighting the importance of timely diagnosis to improve the possibility of treatment with curative intent.12 Additionally, differences in the diagnostic interval for lung cancer has been demonstrated across ICBP countries, supporting the need for greater alignment in international efforts to improve time to diagnosis and treatment parameters.30 Different countries take a more aggressive stance in their time targets with compulsory reporting, and the accountability for these (e.g. Denmark), whilst in others this can be less regimented.31 The network formed in this study have agreed that <30 days from point of referral for possible lung cancer diagnosis should be a minimum requirement to allow best opportunity for timely initiation of treatment. A crucial consideration within this target is the need to ensure timely access to diagnostic and staging tools, with strong coordination between diagnostic and treatment services – the need to improve this from a policy and practice perspective will differ, depending on jurisdictional context.","Developing Thoracic Centres of Excellence and formally affiliated networks can help provide advanced and comprehensive patient-focused approaches to lung cancer treatment. Implementation of this aspiration in different regions may differ but essential components of Centres of Excellence for lung cancer have been researched and agreed by this international network. Expertise from multiple disciplines dedicated to lung cancer management (including thoracic surgery, oncology, anaesthesia, respiratory medicine, pathology, radiology, pulmonology rehabilitation) should be available within the Centre and its network. This enables seamless and multi-disciplinary care for patients. Critical to this is ensuring that robustly trained healthcare professionals and specialist experts relevant to these disciplines are deployed, with investment in their training and development to support continued improvement. Broader formal networks are required to provide contemporary access for patients, alongside streamlined processes across prevention, diagnosis, treatment, and survivorship care. It is important that Centres provide comprehensive care to all patients, including those first referred to local hospitals. There is some evidence in the UK that first referral to a local hospital is associated with fewer patients being offered curative treatment.32 The explanation for this is not clear but may relate to distances that patients need to travel.32 Centres should support availability of clinical trials to help deliver innovative research advances to treatments and to enhance patient outcomes. Ultimately, Thoracic Centres of Excellence and their networks should be viewed as the backbone of timely integrated care for lung cancer patients.18 Development of such Centres has been underway across ICBP countries, but relies heavily on available resource, funding, capacity, human resources, and coordination. When considering service improvement and expansion, Centres of Excellence should be in the forefront of policy plans and considerations.","High-quality data capture and timely feedback is crucial to improving care.33 The importance of this intelligence at local, regional, and national levels needs to be recognised across all countries. Embedding data capture and feedback loops into existing healthcare structures and processes in lung cancer care should be prioritised for self-evaluation and to drive service improvements. Whilst some work has been undertaken to better understand variation in delivery of lung cancer care between comparable high-income countries (as described in the Background), a truly comprehensive understanding of the key drivers of better outcomes in certain countries requires a detailed international audit. In order to deliver this vital intelligence, a number of considerations re required. A universal minimum dataset with key metrics (wait time, stage at diagnosis, risk adjusted resection and radiation rates, medical oncology treatments, significant adverse event rates) is required, alongside strong collaborative relationships between participating countries.34 Ensuring the minimum data metrics are collected as part of routine practice should be the first step to enable a successful audit within jurisdictions, in order to inform the delivery of an international audit.35 Few national audits have been undertaken for lung cancer care, but if more countries replicate this, the roadmap to commissioning an international audit will be clearer.36-38 This will allow clear identification of international best practice to facilitate sharing of optimal care and lessons learnt.","Continuous improvement in lung cancer care and outcomes should be recognised as an expectation by policymakers, requiring ongoing investment. The success of cancer site-specific policies in high-income countries, for example breast cancer, is reflected in enhanced policy focus, driving service development (screening programmes, diagnostic initiatives and treatment options) and underpinning improved survival internationally.39 However, even in higher performing countries, lung cancer survival is still low.2,4 Significant inequities in lung cancer risk, access to care and treatment outcomes have historically existed, driven largely by socioeconomic inequalities, and thus prohibiting improvements in lung cancer survival seen for other cancer sites.40 Recognising and addressing these inequities and inequalities in care collectively needs to be prioritised within cancer policy internationally, facilitating service improvement to optimise management and care of lung cancer patients. In order to drive this improvement, jurisdictions must consider the effectiveness of their leadership (within both clinical and political arenas) to instigate change, prioritise connection to and collaboration with appropriate stakeholders to ensure lung cancer is promoted on political agendas, and have the means to lobby for and disseminate improved lung cancer policy robustly.41 By calling for lung cancer care to be a priority internationally, we aim to encourage policymakers to prioritise the most pressing initiatives required within their respective countries to have the greatest impact upon lung cancer outcomes.","It is acknowledged by the network and the authors that these recommendations were developed in the context of the ICBP country membership and may not fully represent the diversity of healthcare settings globally. We are aware the interviews undertaken to help inform the development of the recommendations were of a limited number, however we targeted these at stakeholders who would have a breadth of knowledge. Data saturation would therefore be unlikely to occur but we hope this format and level of questioning can now be replicated and built upon in other healthcare systems and countries to further develop our understanding of key factors in improving lung cancer care internationally. Both the network and authors hope for the recommendations developed to be applicable to other comparable high-income countries, who can interpret these in the context of their healthcare system structure and funding, cultures, socioeconomic situation, and other population-based needs. Additionally, the importance of rigorously evaluating new initiatives, policies, and the evidence base to support their implementation within local settings should not be underestimated – it is the aim of the network and the authors to bring these recommendations to the forefront of policy attention, in order for local assessment of the requirements needed to evaluate and implement.","The burden of lung cancer is significant and represents a global public health challenge. The Calls to Action and Good Practice Points presented here, reflecting the views of an international expert lung cancer clinical network, signpost the key considerations for policymakers within high-income comparable countries to effective change. They provide a roadmap to help align and focus efforts in improving outcomes and management of lung cancer patients. The Good Practice Points described are key to enable delivery of the Calls to Action, which should be considered the highest priority. Collaboration at local, regional, national, and international levels within the lung cancer community is imperative to empower policymakers to ensure policies are enacted that enhance lung cancer patient treatment and improve quality of life and patient outcomes.","Click here for additional data file.\nSupplemental material for An International Consensus on Actions to Improve Lung Cancer Survival: A Modified Delphi Method Among Clinical Experts in the International Cancer Benchmarking Partnership by Charlotte Lynch, Samantha Harrison, John Butler, David R. Baldwin, Paul Dawkins, Joris van der Horst, Erik Jakobsen, Jonathan McAleese, Annette McWilliams, Karen Redmond, Anand Swaminath, and Christian J. Finley in Cancer Control"],"string":"[\n \"Lung cancer accounts for nearly 1.8 million deaths annually – more than colorectal and breast cancers combined.1 Despite significant improvements in diagnosis and treatment, lung cancer is typically associated with low survival internationally.2 The International Cancer Benchmarking Partnership (ICBP) is a global collaboration of clinicians, policymakers, researchers, and cancer data experts. It seeks to benchmark and explain cancer survival, incidence, and mortality differences between high-income countries with comprehensive cancer registry coverage, similar budgetary spending on national health systems and universal access to healthcare. Recent ICBP research has demonstrated that international disparities in lung cancer survival persist, with 5-year survival ranging from 14.7% (UK) to 21.7% (Canada).2 Evidence shows that 3-year survival varies mostly in potentially curative lung cancer, by up to 20 percentage points for localised and regional disease (as per SEER summary staging)3 across the ICBP countries.4 Comparably, survival varied by just 3.4 percentage points in patients diagnosed with distant/metastatic disease across the same countries.4\\nDue to the inherent complexity of health systems, it is unlikely that one single component of the cancer care continuum can ultimately define why survival variation exists, yet the ability to diagnose and stage lung cancer patients as early as possible will likely influence their treatment options.5 Various work has been undertaken to articulate patient, healthcare professional and disease related factors that influence timeliness of diagnosis and ultimately, patient outcomes.6 This understanding has helped build initiatives across the pathway to address the poorer outcomes in lung cancer, but to differing levels internationally. This starts with preventative initiatives for lung cancer, including support for smoking cessation and lung cancer awareness campaigns. Both areas are of significance, with tobacco being a key factor in the aetiology of lung cancer, and some awareness campaigns influencing stage shift.7,8 Variation in primary care practitioner readiness to refer patients with potential symptoms and in service provision of key diagnostic and staging tools for lung cancer, such as PET-CT, have been demonstrated across ICBP countries with unknown quantifiable impacts upon survival.9,10 This is reflected in the variation seen in lung cancer stage distribution between ICBP countries, but not to the same extent as differences in survival by stage.4 In the diagnostic phase, it is clear that international differences exist for lung cancer patients, yet the evidence for how this affects outcomes is variable and complex.11 Variation within the diagnosis and staging of early stage lung cancer patients is important; unwarranted delays enabling tumour growth may shift patients away from being in a potentially curable stage towards more advanced stages with lower survival.12,13\\nCurative intent treatment options for lung cancer have increased and become significantly more sophisticated over recent years, with the development and adoption of more precise radiotherapy, and minimally invasive surgery.14 Less invasive options have opened up better opportunities for more complex patients or those with greater comorbidities. Guidance and uptake of these treatment options for early stage patients is likely to vary both within countries and internationally. This is particularly the case in patients deemed to be borderline candidates for curative therapy where there are more variables playing a role in decision-making for treatment.15 Variation in the use of curative intent therapies within early stage patients has been demonstrated in some countries.16 Disease stage, patient fitness/performance status, lung function, staffing skillset, socioeconomic status and comorbidities play key influential roles in the perceived operability of patients and subsequent treatment decisions.16\\nThe organisation of each country’s healthcare system has been established to guide patients from presentation to the decision for treatment, with many informal and formal treatment pathways being established. However, gaps and biases within each system have been identified that undoubtedly affect the management of potentially curable patients.17 There is a current lack of internationally agreed recommendations for improving care of lung cancer patients, with the purpose of being interpreted at a policy level to align efforts to empower those in positions to instigate change.\\nWithin this study, we have established an international lung cancer network to share local clinical insights, including locally defined best practices, mitigation of modifiable variables and opportunities for collaboration. We present a series of recommendations to improve lung cancer care, in order to improve survival, based on the consensus of our international clinical network. We hope that these recommendations serve to optimise lung cancer services so as to improve the quality of options for treatment, equitable access to treatment, and ultimately improve survival internationally.\",\n \"[SUBTITLE] Network Formation [SUBSECTION] Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\\nNine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\\n[SUBTITLE] Topic Guide Development and Initial Interviews [SUBSECTION] We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\\nWe initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\\n[SUBTITLE] Roundtable Discussions and Recommendation Development [SUBSECTION] Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\\nFour roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\\n[SUBTITLE] Modified Delphi Method [SUBSECTION] Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\\nStepwise modified Delphi Method.\\nNetwork members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\\nStepwise modified Delphi Method.\\n[SUBTITLE] Refinement of Recommendations [SUBSECTION] ‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.\\n‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.\",\n \"Nine network members were recruited from nine ICBP jurisdictions: Australia, Canada, Denmark, England, Ireland, New Zealand, Northern Ireland, Scotland and Wales. Members were identified via purposive sampling based on them having roles in national, regional, and local tiers of health systems with expertise primarily being thoracic surgery, radiation oncology, clinical oncology, or respiratory medicine. Existing ICBP clinical networks and stakeholders were approached and asked for suggestions for the network, and desk-based research was conducted to assess suitability based on the criteria outlined in the previous sentence. Members were initially invited to join the network via email, and if interested, they were sent a Terms of Reference and invited to a telephone call to discuss the aims of the project and the ICBP in more detail. At that point, members were asked if they agreed to be part of the network and involved in the project.\",\n \"We initially employed a qualitative approach to understand potential contributing factors to international variation in lung cancer outcomes that could form the basis of the recommendations developed. The overarching focus of this was mostly on non-small cell lung cancer (NSCLC), and for early stage lung cancer. Semi-structured interviews were conducted with the network members to identify key themes and points of variation between the ICBP countries. Interview questions (Supplementary Materials) were developed by the lead study team (Finley, Lynch, Butler, Harrison), alongside an evidence scoping exercise and review of published literature. Interviews were recorded via Microsoft Teams and stored securely at Cancer Research UK. Qualitative data was transcribed by the lead study team and interview transcripts were analysed thematically, drawing on documents developed during processes undertaken to improve existing national guidelines and support for improving lung cancer care, largely the Pan-Canadian Standards for Thoracic Surgery.18 The analysis of these interviews helped generate the recommendations.\",\n \"Four roundtable discussions were hosted by the lead study team with all network members. The first and second roundtables discussed the results from the interviews and shared the key themes emerging, which helped structure the discussion to begin generating the recommendations. Network members agreed on key topics for the recommendations to be based on (e.g. pre-diagnosis and screening), and the lead study team articulated these into written recommendations. The third and fourth roundtable meetings were hosted to receive feedback on the Delphi rating method (described below) and further refine the recommendations.\",\n \"Network members were asked to rate recommendations using a modified Delphi consensus model (Figure 1).19 In the first round, interviewees were asked to rate 22 recommendations developed from the interview results and initial suggestions from the first and second roundtable discussions with the network members. Options to ‘agree’, ‘disagree’ and provide additional comments and/or suggestions within a free text box were available. A minimum satisfactory consensus of 70% was reached in the first round, with agreement on all recommendations. Subsequent roundtable discussions were hosted to obtain further feedback and consolidate agreement on any recommendations disagreed with or requiring modification. In the next round, 20 revised recommendations were shared with the same participants to rate agreement with the same format (agree, disagree, additional comments/suggestions). The recommendations were then further refined, with a final round of consensus voting resulting in 18 final recommendations.Figure 1.Stepwise modified Delphi Method.\\nStepwise modified Delphi Method.\",\n \"‘Calls to Action’ for each phase of care were identified by the network as the most crucial recommendations via Delphi consensus voting, requiring the most attention by and focus of policymakers, with ‘Good Practice Points’ being required to support the implementation of these Calls to Action. Good Practice points were further categorised broadly into phases of care (pre-diagnosis, diagnosis, therapeutic and all phases) for ease of implementation.\",\n \"[SUBTITLE] Initial Interview Results [SUBSECTION] These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\\nThese are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\\n[SUBTITLE] Recommendations [SUBSECTION] A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\\nA total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\\n[SUBTITLE] Calls to Action Summary [SUBSECTION] Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nCalls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\n[SUBTITLE] Good Practice Points Summary [SUBSECTION] The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\nThe network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"These are reported in the Supplementary Materials. Key themes in the differences reported between countries include: variation in data capture and auditing, composition and use of multidisciplinary teams, access to mitigating programmes for vulnerable populations and access to clinical research.\\nThe remainder of Results section focuses on the recommendations developed during this study as these are our core output.\",\n \"A total of 18 recommendations achieved 100% consensus. These were further grouped in to five Calls to Action (Box 1A) and thirteen Good Practice Points (Box 1B).\",\n \"Calls to Action should be interpreted as key priorities to inform policy not only across the ICBP countries, but in similar high-income countries globally. It is recognised that different countries will have differing resource, capacity, funding, and population-based needs and should action these recommendations accordingly to their local settings.Box 1A: Calls to ActionImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\n[SUBTITLE] Box 1A: Calls to Action [SUBSECTION] Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\",\n \"Implement cost-effective, clinically efficacious, and equitable lung cancer screening initiativesEnsure diagnosis of lung cancer within 30 days of referralDevelop Thoracic Centres of ExcellenceUndertake an international audit of lung cancer careRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\\nImplement cost-effective, clinically efficacious, and equitable lung cancer screening initiatives\\nEnsure diagnosis of lung cancer within 30 days of referral\\nDevelop Thoracic Centres of Excellence\\nUndertake an international audit of lung cancer care\\nRecognise improvements in lung cancer care and outcomes as a priority in cancer policy\",\n \"The network agreed these points are important considerations for similar high-income countries in optimising management and treatment of lung cancer patients. They should be considered as key in supporting the development and implementation of the Calls to Action.Box 1B: Good Practice PointsPre-Diagnosis• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.Diagnosis• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.Therapeutic• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.All Phases of Care• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Box 1B: Good Practice Points [SUBSECTION] [SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"[SUBTITLE] Pre-Diagnosis [SUBSECTION] • Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n[SUBTITLE] Diagnosis [SUBSECTION] • Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n[SUBTITLE] Therapeutic [SUBSECTION] • Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n[SUBTITLE] All Phases of Care [SUBSECTION] • Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\\n• Timely and widespread availability for primary care referrals to cross sectional imaging should be prioritised in patients with a suspicion of lung cancer.\\n• National coordinated education materials and support should be available to public and health care providers on lung cancer and its management and outcomes.\",\n \"• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\\n• Where appropriate, lung malignancies should be discussed in a multidisciplinary format with essential components in place:• Protocols for incorporation into workplans, coordination and chairing of meetings• Appropriate attendance from core stakeholders• Means to record and report treatment decisions locally and centrally• Method to communicate treatment decisions to referring practitioners\\n• Protocols for incorporation into workplans, coordination and chairing of meetings\\n• Appropriate attendance from core stakeholders\\n• Means to record and report treatment decisions locally and centrally\\n• Method to communicate treatment decisions to referring practitioners\\n• Molecular profiling of tumours should be prioritised where there is a clinical need, with pathology reports presented in a synoptic format and communicated within 2 weeks of biopsy or operation.\",\n \"• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\\n• Incorporation of pre-habilitation programmes into routine care for patients undergoing surgical resection should be considered and implemented where appropriate.\\n• Curative intent therapy should be prioritised where possible, even in higher risk patients where the challenges in management of significant comorbidities and increased risk of adverse events is recognised.\\n• Integrated care teams should be in place and supported by surgeons with sufficient topical expertise to identify preventable adverse events in patients.\\n• It is the expectation that when adopting new technologies and techniques active tracking of adverse events and outcomes will be completed.\",\n \"• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\\n• Each country should have in place a minimum dataset for the evaluation of lung cancer patients’ diagnosis, treatment, and aftercare. Minimum datasets should have at least the following critical data elements: wait times, stage at diagnosis, risk adjusted resection and radiation rates, and significant adverse event rates.\\n• Commissioning of regular national clinical audits should be employed in all countries, with necessary analytical capacity for timely analysis and interrogation of data at local levels\\n• Emphasis on quality improvement should be placed in every part of the lung cancer pathway with regular review of audit data and processes in place to facilitate sharing and learning of best practice.\\n• Identification of patients at high-risk for negative outcomes should be part of routine care, with the development of appropriate pathways to mitigate poorer outcomes.\\n• System and investment planning for future advanced equipment should be undertaken to reflect emerging/new and innovative diagnostic methods and therapeutics.\",\n \"Rapid progress in areas of healthcare need can be achieved when international collaborative efforts, research and funding are aligned to a common goal, most recently illustrated by the development of the COVID vaccine.20 The international cancer community recognises the need to significantly improve lung cancer survival.21 The Calls to Action and Good Practice Points generated in this study provide a baseline for reflection, appropriate benchmarking, and identification of opportunities for enhancing lung cancer care. These can be interpreted from local, regional, national, and international perspectives. Many advances have been made in lung cancer prevention, screening, diagnosis, and treatment worldwide – from lower smoking rates, utilisation of CT scans for earlier detection, greater access to innovative medicines, improved operative techniques and more advanced radiation therapy. However, the potential improvements in survival have not yet been achieved. The international network formed, and recommendations developed within this study, are advocating for increased and better alignment of efforts within, and across, countries in order to improve lung cancer care and survival. We outline the importance and considerations to be had when addressing the Calls to Action below.\",\n \"Due to variation seen in stage distribution between ICBP countries, it is important to consider where opportunities exist to optimise care and improve survival in the pre-treatment interval. Whilst the evidence is strong for a survival benefit of low dose CT screening in individuals with higher risk of lung cancer, implementation has been a complex issue for policymakers.22-24 Whilst low dose CT screening can support detection of early stage disease and increase the likelihood of curable treatment, several factors must be carefully considered to ensure clinical efficacy and cost effectiveness. Integration into existing infrastructure, target populations, recruitment of population at risk, reduction of harm, resource and cost, and engagement with the public and healthcare professionals need to be defined and implemented. Screening programmes need to be designed to reach a traditionally ‘hard to reach’ population, requiring their design to have greater engagement of those affected. There is no “one size fits all” approach, yet it is the collective opinion of this international clinical network that understanding the impact that screening could have for outcomes and considering all these factors within local contexts is a crucial activity for policymakers. Currently, no ICBP country has fully implemented a national lung cancer screening programme, though a number of countries are at different stages of exploration and implementation. Successful models of assessment and pilot trials have been underway internationally for several years and can serve as a baseline to inform the best implementation plan for screening in each jurisdiction.25-29\",\n \"Various temporal targets for key time points within the cancer patient journey exist for different countries, but there has previously been a lack of international consensus on a universal ‘best practice’ target. Unwarranted delays in diagnosis can rapidly shift lung cancer patients from being potentially curative to being difficult to treat, highlighting the importance of timely diagnosis to improve the possibility of treatment with curative intent.12 Additionally, differences in the diagnostic interval for lung cancer has been demonstrated across ICBP countries, supporting the need for greater alignment in international efforts to improve time to diagnosis and treatment parameters.30 Different countries take a more aggressive stance in their time targets with compulsory reporting, and the accountability for these (e.g. Denmark), whilst in others this can be less regimented.31 The network formed in this study have agreed that <30 days from point of referral for possible lung cancer diagnosis should be a minimum requirement to allow best opportunity for timely initiation of treatment. A crucial consideration within this target is the need to ensure timely access to diagnostic and staging tools, with strong coordination between diagnostic and treatment services – the need to improve this from a policy and practice perspective will differ, depending on jurisdictional context.\",\n \"Developing Thoracic Centres of Excellence and formally affiliated networks can help provide advanced and comprehensive patient-focused approaches to lung cancer treatment. Implementation of this aspiration in different regions may differ but essential components of Centres of Excellence for lung cancer have been researched and agreed by this international network. Expertise from multiple disciplines dedicated to lung cancer management (including thoracic surgery, oncology, anaesthesia, respiratory medicine, pathology, radiology, pulmonology rehabilitation) should be available within the Centre and its network. This enables seamless and multi-disciplinary care for patients. Critical to this is ensuring that robustly trained healthcare professionals and specialist experts relevant to these disciplines are deployed, with investment in their training and development to support continued improvement. Broader formal networks are required to provide contemporary access for patients, alongside streamlined processes across prevention, diagnosis, treatment, and survivorship care. It is important that Centres provide comprehensive care to all patients, including those first referred to local hospitals. There is some evidence in the UK that first referral to a local hospital is associated with fewer patients being offered curative treatment.32 The explanation for this is not clear but may relate to distances that patients need to travel.32 Centres should support availability of clinical trials to help deliver innovative research advances to treatments and to enhance patient outcomes. Ultimately, Thoracic Centres of Excellence and their networks should be viewed as the backbone of timely integrated care for lung cancer patients.18 Development of such Centres has been underway across ICBP countries, but relies heavily on available resource, funding, capacity, human resources, and coordination. When considering service improvement and expansion, Centres of Excellence should be in the forefront of policy plans and considerations.\",\n \"High-quality data capture and timely feedback is crucial to improving care.33 The importance of this intelligence at local, regional, and national levels needs to be recognised across all countries. Embedding data capture and feedback loops into existing healthcare structures and processes in lung cancer care should be prioritised for self-evaluation and to drive service improvements. Whilst some work has been undertaken to better understand variation in delivery of lung cancer care between comparable high-income countries (as described in the Background), a truly comprehensive understanding of the key drivers of better outcomes in certain countries requires a detailed international audit. In order to deliver this vital intelligence, a number of considerations re required. A universal minimum dataset with key metrics (wait time, stage at diagnosis, risk adjusted resection and radiation rates, medical oncology treatments, significant adverse event rates) is required, alongside strong collaborative relationships between participating countries.34 Ensuring the minimum data metrics are collected as part of routine practice should be the first step to enable a successful audit within jurisdictions, in order to inform the delivery of an international audit.35 Few national audits have been undertaken for lung cancer care, but if more countries replicate this, the roadmap to commissioning an international audit will be clearer.36-38 This will allow clear identification of international best practice to facilitate sharing of optimal care and lessons learnt.\",\n \"Continuous improvement in lung cancer care and outcomes should be recognised as an expectation by policymakers, requiring ongoing investment. The success of cancer site-specific policies in high-income countries, for example breast cancer, is reflected in enhanced policy focus, driving service development (screening programmes, diagnostic initiatives and treatment options) and underpinning improved survival internationally.39 However, even in higher performing countries, lung cancer survival is still low.2,4 Significant inequities in lung cancer risk, access to care and treatment outcomes have historically existed, driven largely by socioeconomic inequalities, and thus prohibiting improvements in lung cancer survival seen for other cancer sites.40 Recognising and addressing these inequities and inequalities in care collectively needs to be prioritised within cancer policy internationally, facilitating service improvement to optimise management and care of lung cancer patients. In order to drive this improvement, jurisdictions must consider the effectiveness of their leadership (within both clinical and political arenas) to instigate change, prioritise connection to and collaboration with appropriate stakeholders to ensure lung cancer is promoted on political agendas, and have the means to lobby for and disseminate improved lung cancer policy robustly.41 By calling for lung cancer care to be a priority internationally, we aim to encourage policymakers to prioritise the most pressing initiatives required within their respective countries to have the greatest impact upon lung cancer outcomes.\",\n \"It is acknowledged by the network and the authors that these recommendations were developed in the context of the ICBP country membership and may not fully represent the diversity of healthcare settings globally. We are aware the interviews undertaken to help inform the development of the recommendations were of a limited number, however we targeted these at stakeholders who would have a breadth of knowledge. Data saturation would therefore be unlikely to occur but we hope this format and level of questioning can now be replicated and built upon in other healthcare systems and countries to further develop our understanding of key factors in improving lung cancer care internationally. Both the network and authors hope for the recommendations developed to be applicable to other comparable high-income countries, who can interpret these in the context of their healthcare system structure and funding, cultures, socioeconomic situation, and other population-based needs. Additionally, the importance of rigorously evaluating new initiatives, policies, and the evidence base to support their implementation within local settings should not be underestimated – it is the aim of the network and the authors to bring these recommendations to the forefront of policy attention, in order for local assessment of the requirements needed to evaluate and implement.\",\n \"The burden of lung cancer is significant and represents a global public health challenge. The Calls to Action and Good Practice Points presented here, reflecting the views of an international expert lung cancer clinical network, signpost the key considerations for policymakers within high-income comparable countries to effective change. They provide a roadmap to help align and focus efforts in improving outcomes and management of lung cancer patients. The Good Practice Points described are key to enable delivery of the Calls to Action, which should be considered the highest priority. Collaboration at local, regional, national, and international levels within the lung cancer community is imperative to empower policymakers to ensure policies are enacted that enhance lung cancer patient treatment and improve quality of life and patient outcomes.\",\n \"Click here for additional data file.\\nSupplemental material for An International Consensus on Actions to Improve Lung Cancer Survival: A Modified Delphi Method Among Clinical Experts in the International Cancer Benchmarking Partnership by Charlotte Lynch, Samantha Harrison, John Butler, David R. Baldwin, Paul Dawkins, Joris van der Horst, Erik Jakobsen, Jonathan McAleese, Annette McWilliams, Karen Redmond, Anand Swaminath, and Christian J. Finley in Cancer Control\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["intro","materials|methods",null,null,null,null,null,"results",null,null,null,null,null,null,null,null,null,null,"discussion",null,null,null,null,null,null,"conclusion","supplementary-material"],"string":"[\n \"intro\",\n \"materials|methods\",\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"conclusion\",\n \"supplementary-material\"\n]"},"keywords":{"kind":"list like","value":["lung cancer","health policy","public health","epidemiology","survival"],"string":"[\n \"lung cancer\",\n \"health policy\",\n \"public health\",\n \"epidemiology\",\n \"survival\"\n]"}}},{"rowIdx":387,"cells":{"title":{"kind":"string","value":"Disinfection of Virtual Reality Devices in Health Care Settings: In Vitro Assessment and Survey Study."},"pmid":{"kind":"string","value":"36269222"},"background_abstract":{"kind":"string","value":"Virtual reality (VR) devices are increasingly used in health care settings. The use among patients has the potential to unintentionally transmit pathogens between patients and hospital staff. No standard operating procedure for disinfection exists to ensure safe use between patients."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"Three types of bacteria were inoculated onto porous and nonporous surfaces of 2 VR devices: the Meta Oculus Quest and Meta Oculus Quest 2. Disinfection was performed using either isopropyl alcohol or alcohol-free quaternary ammonium wipes. A quantitative culture was used to assess the adequacy of disinfection. A survey was separately sent out to VR device technicians at other pediatric health care institutes to compare the methods of disinfection and how they were established."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Both products achieved adequate disinfection of the treated surfaces; however, a greater log-kill was achieved on nonporous surfaces than on the porous surfaces. Alcohol performed better than quaternary ammonium on porous surfaces. The survey respondents reported a wide variability in disinfection processes with only 1 person reporting an established standard operating procedure."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Disinfection can be achieved through the use of either isopropyl alcohol or quaternary ammonium products. Porous surfaces showed lesser log-kill rates than the nonporous surfaces, indicating that the use of an added barrier may be of benefit and should be a point of future research. Given the variability in the disinfection process across health care systems, a standard operating procedure is proposed."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSIONS"},"mesh_descriptor_names":{"kind":"list like","value":["Child","Humans","Disinfection","2-Propanol","Ethanol","Ammonium Compounds","Surveys and Questionnaires","Delivery of Health Care","Virtual Reality"],"string":"[\n \"Child\",\n \"Humans\",\n \"Disinfection\",\n \"2-Propanol\",\n \"Ethanol\",\n \"Ammonium Compounds\",\n \"Surveys and Questionnaires\",\n \"Delivery of Health Care\",\n \"Virtual Reality\"\n]"},"pmcid":{"kind":"string","value":"9756115"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"Virtual reality (VR) devices are increasingly used in health care settings to benefit patients, and the examples include patients with posttraumatic stress disorder, anxiety, complex regional pain syndrome, and distraction therapy [1-4]. Recent data show the benefit expands to the pediatric population as well by reducing pain and anxiety during medical procedures through distraction [5]. VR can also be used to educate health care workers through training and simulation [6]. However, a lack of standardized cleaning and disinfection processes for VR devices has limited VR’s use in health care settings, especially during the COVID-19 pandemic [6].\nNosocomial transmission and outbreaks have been reported with many different types of medical devices in clinical use [7], and establishing standard operating procedures (SOPs) for disinfection of VR devices between patient use is paramount. One of the most common and widely sold VR headset devices worldwide, the Meta Oculus Quest 2, specifically, recommends against the use of alcohol to clean and disinfect the device in favor of antibacterial wipes due to theoretical concerns about affecting the porous material [8]. This poses challenges in clinical settings as isopropyl alcohol (IPA) is one of the most common disinfectants used on medical devices.\nA protocol to clean and disinfect VR devices used in health care settings has been proposed [6]; however, no studies have quantified the efficacy of hospital-grade disinfectants on different parts of the VR equipment. Additionally, little is known about how these machines are currently disinfected in health care settings. In this mixed methods evaluation, we sought to determine the current disinfection practices in health care settings and how they were established. We also studied the effect of commonly used disinfectant wipes on the disinfection of VR headsets experimentally contaminated with common bacterial pathogens to provide evidence for the creation of an SOP to reduce infections with multipatient VR utilization."},"methods_title":{"kind":"string","value":"Methods"},"methods_text":{"kind":"string","value":"[SUBTITLE] Survey [SUBSECTION] To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\nTo learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\n[SUBTITLE] Ethical Considerations [SUBSECTION] The survey was approved as exempt by the Yale University institutional review board (study #2000033075).\nThe survey was approved as exempt by the Yale University institutional review board (study #2000033075).\n[SUBTITLE] Laboratory Disinfection [SUBSECTION] Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\nThree types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\n[SUBTITLE] Statistical Analysis [SUBSECTION] The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\nThe experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05."},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Current VR Disinfection Practices in Pediatric Hospitals [SUBSECTION] A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\nA total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\n[SUBTITLE] Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination [SUBSECTION] A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nA total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Survey","Ethical Considerations","Laboratory Disinfection","Statistical Analysis","Current VR Disinfection Practices in Pediatric Hospitals","Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination"],"string":"[\n \"Survey\",\n \"Ethical Considerations\",\n \"Laboratory Disinfection\",\n \"Statistical Analysis\",\n \"Current VR Disinfection Practices in Pediatric Hospitals\",\n \"Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination\"\n]"},"other_sections_texts":{"kind":"list like","value":["To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.","The survey was approved as exempt by the Yale University institutional review board (study #2000033075).","Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.","The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.","A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.","A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap."],"string":"[\n \"To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\",\n \"The survey was approved as exempt by the Yale University institutional review board (study #2000033075).\",\n \"Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\",\n \"The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\",\n \"A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\\nSurvey of current VRa disinfection practices in health care settings.\\naVR: virtual reality.\\nbSome respondents gave multiple responses.\\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\",\n \"A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\\nBacterial count by disinfection method, organism, and surface.\\naThe percentage of times when 0 bacterial colony-forming units were observed.\\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Introduction","Methods","Survey","Ethical Considerations","Laboratory Disinfection","Statistical Analysis","Results","Current VR Disinfection Practices in Pediatric Hospitals","Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination","Discussion"],"string":"[\n \"Introduction\",\n \"Methods\",\n \"Survey\",\n \"Ethical Considerations\",\n \"Laboratory Disinfection\",\n \"Statistical Analysis\",\n \"Results\",\n \"Current VR Disinfection Practices in Pediatric Hospitals\",\n \"Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination\",\n \"Discussion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Virtual reality (VR) devices are increasingly used in health care settings to benefit patients, and the examples include patients with posttraumatic stress disorder, anxiety, complex regional pain syndrome, and distraction therapy [1-4]. Recent data show the benefit expands to the pediatric population as well by reducing pain and anxiety during medical procedures through distraction [5]. VR can also be used to educate health care workers through training and simulation [6]. However, a lack of standardized cleaning and disinfection processes for VR devices has limited VR’s use in health care settings, especially during the COVID-19 pandemic [6].\nNosocomial transmission and outbreaks have been reported with many different types of medical devices in clinical use [7], and establishing standard operating procedures (SOPs) for disinfection of VR devices between patient use is paramount. One of the most common and widely sold VR headset devices worldwide, the Meta Oculus Quest 2, specifically, recommends against the use of alcohol to clean and disinfect the device in favor of antibacterial wipes due to theoretical concerns about affecting the porous material [8]. This poses challenges in clinical settings as isopropyl alcohol (IPA) is one of the most common disinfectants used on medical devices.\nA protocol to clean and disinfect VR devices used in health care settings has been proposed [6]; however, no studies have quantified the efficacy of hospital-grade disinfectants on different parts of the VR equipment. Additionally, little is known about how these machines are currently disinfected in health care settings. In this mixed methods evaluation, we sought to determine the current disinfection practices in health care settings and how they were established. We also studied the effect of commonly used disinfectant wipes on the disinfection of VR headsets experimentally contaminated with common bacterial pathogens to provide evidence for the creation of an SOP to reduce infections with multipatient VR utilization.","[SUBTITLE] Survey [SUBSECTION] To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\nTo learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\n[SUBTITLE] Ethical Considerations [SUBSECTION] The survey was approved as exempt by the Yale University institutional review board (study #2000033075).\nThe survey was approved as exempt by the Yale University institutional review board (study #2000033075).\n[SUBTITLE] Laboratory Disinfection [SUBSECTION] Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\nThree types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\n[SUBTITLE] Statistical Analysis [SUBSECTION] The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\nThe experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.","To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.","The survey was approved as exempt by the Yale University institutional review board (study #2000033075).","Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.","The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.","[SUBTITLE] Current VR Disinfection Practices in Pediatric Hospitals [SUBSECTION] A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\nA total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\n[SUBTITLE] Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination [SUBSECTION] A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nA total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.","A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\nSurvey of current VRa disinfection practices in health care settings.\naVR: virtual reality.\nbSome respondents gave multiple responses.\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.","A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\nBacterial count by disinfection method, organism, and surface.\naThe percentage of times when 0 bacterial colony-forming units were observed.\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.","Adequate disinfection of VR devices can be achieved through the use of low-level EPA-approved hospital-grade disinfectants commonly used in clinical settings for devices that are exposed to intact skin, including IPA and quaternary ammonium wipes. We found a greater than 6-fold logarithmic reduction from initial bacteria inoculation across all pathogen types and VR device surfaces when using either product. However, we did observe the differences when evaluating raw-bacterial counts after disinfection. Notably, IPA performed better than the quaternary ammonium wipe, particularly for porous surfaces. It is possible that IPA penetrates porous surfaces better than quaternary ammonium products due to the vigorous 15-second scrub, and future studies should evaluate how well different wipes perform on these types of surfaces. We also observed that S aureus and S epidermidis persisted on surfaces at greater densities than P aeruginosa, possibly reflective of a mechanism in strain type or environmental survivability. Finally, and perhaps most critically, we observed lower bacterial counts after inoculation but before disinfection, and greater bacterial counts after disinfection, on porous surfaces when compared to nonporous surfaces. This suggests that bacteria may be entering the pores in the material, potentially reducing exposure to the disinfection material. Additionally, using swabs to recover bacteria from porous surfaces is suboptimal as we do not recover bacteria that have penetrated deeper into the material as well as nonporous surfaces which have better transfer efficiency [10]. Thus, despite consistent recovery after disinfection from porous surfaces in these experiments, we likely have overestimated the efficacy of disinfection for this material. Of note, there was outstanding disinfection of all nonporous surfaces, making it the preferred material for the construction of VR devices in health care. Manufacturers should consider material in the design of both headsets and straps, and our data support the use of nonporous material, particularly in health care settings where persistent bacteria may serve as a nidus for transmission to the next VR device user. If porous surfaces are present, there should be adequate barrier protection to prevent the transmission of microbes. This also then allows for the use of IPA without concern for damage to any porous components of the device.\nNosocomial transmission and outbreaks associated with the use of medical devices are well documented and a primary concern for using VR devices in health care settings [7]. Consequently, facilities may restrict VR devices from patient use out of concern. We found substantial variability between facilities in the frequency of device use, disinfection method, and barrier protection used. Importantly, almost all sites reported that infection prevention teams were not involved in performing a risk assessment for device use during patient care, and SOPs for disinfection were absent in all but one institution. Establishing a standard process that appropriately disinfects VR devices to allow for safe and expanded use in health care settings while avoiding equipment degradation can benefit patients and health care workers alike. It is critical to ensure that when new devices such as VR equipment are introduced into patient care settings, Infection Prevention and other stakeholders are involved prior to the purchase of the devices to ensure there is an acceptable plan for device reprocessing.\nBased on this generated data set, manufacturer’s instructions for use [8], health care infection prevention best practices, and previous literature or expert opinion [6], we propose an SOP for VR use and disinfection in the health care setting (Textbox 1). This is particularly important as the patient population served may be undergoing chemotherapy or other immunosuppressants that can increase the risk of infection.\n\nBefore use\n\nAvoid on patients with nonintact skin or active infections on the head or hands that cannot be covered and might come into contact with the device\nAvoid use on patients known to be colonized with pathogens where specialized disinfection is required, including Clostridioides difficile, Candida auris, Mycobacterium tuberculosis, and nonenveloped viruses\nPatient and staff perform hand hygiene\nA nonporous cover over the face pad, a disposable face pad cover, or both should serve as a barrier between the patient’s face and the device. Hair should also be covered (eg, bouffant and washable cloth surgical cap). Any porous material that makes contact with the patient’s skin or hair should be covered with a barrier\nDevices should be assessed for alcohol compatibility. If the device is not alcohol compatible, a nonalcohol-based disinfectant should be used\nPerform disinfection with a device compatible Environmental Protection Agency–registered product List H [11] according to the manufacturer’s instructions for use, ensuring that all surfaces of the headset (including the strap, the casing, the inner and outer facepieces, and the lens), the controller, and the nonporous, nondisposable face cover are saturated. Do not use wipes on multiple devices.\n\nAfter use\n\nPatient and staff perform hand hygiene\nStaff don appropriate personal protective equipment, which should include nitrile gloves at a minimum unless other personal protective equipment is required per the patient’s transmission-based isolation precautions\nRemove the device from the patient and placed on a clean disposable pad\nDiscard the disposable face cover, if present\nRemove nonporous, nondisposable face cover from the device, if present\nClean all visibly soiled areas with disposable wipes or paper towels\nRepeat disinfection as above\nAllow headset and controllers to dry according to the product instructions for use\nStore the device in a dry space physically separated from nondisinfected devices\nPatient and staff perform hand hygiene\nIn creating the SOP (Textbox 1), we considered VR devices a noncritical item requiring low-level disinfection between patients because they are most commonly exposed to intact skin only. Given disinfection success with the IPA and alcohol-free quaternary ammonium wipes, we suspect other equivalent low-level disinfectant products (eg, combination IPA or quaternary ammonium wipes) would be adequate, especially when applied to nonporous services [12]. We did not evaluate high-level disinfectants or sterilization procedures that may be required in the event of device exposure to nonintact skin or mucous membranes. We suggest avoiding VR device use on patients who have breaks in the skin on the hands or head region that cannot be appropriately covered and could come into contact with the device, thus avoiding the need for a high level of disinfection.\nAs disinfection was successfully achieved for a variety of pathogens, VR use is most likely safe for patients where contact isolation (gowns and gloves) is required in the hospital, including patients colonized with methicillin-resistant S aureus or vancomycin-resistant Enterococci. However, we suggest clinicians exercise caution when using VR devices with patients colonized with harder-to-eradicate pathogens such as Clostridioides difficile, Candida auris, and nonenveloped viruses where sodium hypochlorite or other high-level disinfection methods may be required. The pathogens tested (S aureus, P aeruginosa, and S epidermidis) are very common organisms in health care settings seen in both the adult and pediatric populations, and these results from disinfection are likely to be applicable to most health care settings, regardless of the patient’s age.\nLimitations of this study include the single-site nature limiting generalizability and the poor survey response rate. We also only tested 2 VR devices from the same company and 2 methods of surface disinfection. Preliminary experiments with an ultraviolet C (UVC) device specifically designed to decontaminate VR devices failed to produce adequate disinfection (data not shown). Since UVC disinfection depends on the angle and distance from the surface to the UVC source [13], the geometry of the headsets may make UVC disinfection more challenging. Further studies evaluating alternative disinfection methods, including UVC and other types of VR devices, are ongoing."],"string":"[\n \"Virtual reality (VR) devices are increasingly used in health care settings to benefit patients, and the examples include patients with posttraumatic stress disorder, anxiety, complex regional pain syndrome, and distraction therapy [1-4]. Recent data show the benefit expands to the pediatric population as well by reducing pain and anxiety during medical procedures through distraction [5]. VR can also be used to educate health care workers through training and simulation [6]. However, a lack of standardized cleaning and disinfection processes for VR devices has limited VR’s use in health care settings, especially during the COVID-19 pandemic [6].\\nNosocomial transmission and outbreaks have been reported with many different types of medical devices in clinical use [7], and establishing standard operating procedures (SOPs) for disinfection of VR devices between patient use is paramount. One of the most common and widely sold VR headset devices worldwide, the Meta Oculus Quest 2, specifically, recommends against the use of alcohol to clean and disinfect the device in favor of antibacterial wipes due to theoretical concerns about affecting the porous material [8]. This poses challenges in clinical settings as isopropyl alcohol (IPA) is one of the most common disinfectants used on medical devices.\\nA protocol to clean and disinfect VR devices used in health care settings has been proposed [6]; however, no studies have quantified the efficacy of hospital-grade disinfectants on different parts of the VR equipment. Additionally, little is known about how these machines are currently disinfected in health care settings. In this mixed methods evaluation, we sought to determine the current disinfection practices in health care settings and how they were established. We also studied the effect of commonly used disinfectant wipes on the disinfection of VR headsets experimentally contaminated with common bacterial pathogens to provide evidence for the creation of an SOP to reduce infections with multipatient VR utilization.\",\n \"[SUBTITLE] Survey [SUBSECTION] To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\\nTo learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\\n[SUBTITLE] Ethical Considerations [SUBSECTION] The survey was approved as exempt by the Yale University institutional review board (study #2000033075).\\nThe survey was approved as exempt by the Yale University institutional review board (study #2000033075).\\n[SUBTITLE] Laboratory Disinfection [SUBSECTION] Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\\nThree types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\\n[SUBTITLE] Statistical Analysis [SUBSECTION] The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\\nThe experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\",\n \"To learn how health care facilities disinfect VR equipment and whether infection prevention teams are involved, a voluntary Qualtrics survey was sent via an electronic link in a group chat of 50 VR technicians working at children’s hospitals across the United States as a convenience sample.\",\n \"The survey was approved as exempt by the Yale University institutional review board (study #2000033075).\",\n \"Three types of bacteria, Staphylococcus epidermidis (ATCC 12228), Pseudomonas aeruginosa (laboratory strain PAO1), and Staphylococcus aureus (ATCC 25923), were chosen because of their propensity to be present on the skin and cause infection in children with compromised immune systems. The bacteria were grown overnight in 3 mL of lysogeny broth and serially diluted to quantitate the bacterial inoculum. VR headsets and controllers were inoculated by spreading 10 μL (initial inoculum 4.1×106-4.5×108) onto various sites (Figure 1) and allowed to dry for 30 minutes. This large inoculum was chosen to test whether Environmental Protection Agency (EPA)–approved disinfectants achieved sufficient log-reduction in bacteria as per their instructions for use. Two VR devices were experimentally contaminated: the Oculus Quest headset and the Oculus Quest 2 headset and hand controllers (Reality Labs, Meta Platforms). These devices were chosen as these are the most popular consumer devices, the primary ones used at our institution, and contained different surface types to trial disinfection [9]. Contaminated sites included the outer surface of the headset housing on the top side, the controller buttons, and the headband straps for each device (Figure 1). These sites were chosen as they were thought to be high–touch point areas for the hands and head during patient use. We did not study the facial interface as it is our standard practice to use a disposable barrier between the facial interface and the patient’s skin. Both nonporous (Oculus Quest 1 strap, Oculus Quest 2 headset, and controller) and porous (Oculus Quest 1 headset and Oculus Quest 2 strap) surfaces were contaminated to assess disinfection efficacy.\\nTwo products were tested for active disinfection: a 70% IPA wipe (Medium Alcohol Prep Pad, Medline) and an alcohol-free quaternary ammonium wipe (Sani-Cloth AF3 Germicidal Disposable Wipe, PDI). A positive control of inoculation without disinfection was performed for every experiment and cultured after the dry time to account for bacterial cell death from desiccation. Disinfection was performed in accordance with each product’s manufacturer’s instructions for use. For the IPA wipe, a 15-second scrub in a back-and-forth motion followed by a 15-second dry time was performed. For the alcohol-free quaternary ammonium wipe, the experimentally contaminated area was wiped to a point of saturation for 3 minutes and allowed to dry. Following disinfection, the cultures were obtained with sterile cotton swabs dipped in Dey-Engley (D/E) neutralizing broth (Hardy Diagnostics) and wiped across the entirety of the contaminated surface for 5 seconds in a back-and-forth motion. The swab was used to inoculate a D/E agar plate that was then incubated overnight at 37 C. Bacterial colony forming units (CFUs) were counted on the plates the following day. Between experiments, the entirety of the VR devices was disinfected with a 70% IPA spray to the point of saturation of the surface materials and dried overnight. The laboratory experiments satisfied the Yale University 100 CH.9 Clinical Quality Improvement criteria and were exempt from institutional review board approval.\\nAreas where cultures were obtained (shown by red circles) of Oculus Quest (top) and Oculus Quest 2 (bottom) devices.\",\n \"The experimental design was a 3-factor crossed design with bacterial CFUs as the outcome variable. The generalized linear model with binomial distribution and logit link was used to compare whether the proportion of trials with observable bacteria counts after disinfection differed by (1) type of disinfection, (2) type of organism, and (3) surface type. All models included the natural log of bacterial count prior to disinfection and used robust standard errors. Achieving disinfection was defined as a greater than a 6-log reduction in bacterial counts. Pairwise comparisons were performed for the type of organism. Analyses were performed in SPSS (version 27; IBM Corp), and statistical significance was set at an α level of .05.\",\n \"[SUBTITLE] Current VR Disinfection Practices in Pediatric Hospitals [SUBSECTION] A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\\nSurvey of current VRa disinfection practices in health care settings.\\naVR: virtual reality.\\nbSome respondents gave multiple responses.\\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\\nA total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\\nSurvey of current VRa disinfection practices in health care settings.\\naVR: virtual reality.\\nbSome respondents gave multiple responses.\\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\\n[SUBTITLE] Effectiveness of Hospital-Grade Disinfecting Wipes on VR Decontamination [SUBSECTION] A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\\nBacterial count by disinfection method, organism, and surface.\\naThe percentage of times when 0 bacterial colony-forming units were observed.\\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\\nA total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\\nBacterial count by disinfection method, organism, and surface.\\naThe percentage of times when 0 bacterial colony-forming units were observed.\\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\",\n \"A total of 50 VR technicians across the Unites States were invited to participate in the Qualtrics survey with a response rate of 18% (9/50). One person consented and then did not answer any of the questions. The selected results of the survey are shown in Table 1 and highlight the variability of VR use and disinfection practices.\\nSurvey of current VRa disinfection practices in health care settings.\\naVR: virtual reality.\\nbSome respondents gave multiple responses.\\nThe number of VR sessions varied from less than once a week to multiple times per day. Most locations used multiple types of VR platforms with the Starlight Children’s VR system most commonly used. The Starlight Children’s VR system is a variation of the Lenovo Mirage Solo VR headset and is made of very similar materials as the Oculus Quest 1 and Oculus Quest 2. All participants noted that disinfection was performed before patient use (n=9, 100%). The methods of disinfection were variable with IPA and quaternary ammonium low-level disinfection wipes used most commonly. Most VR technicians used physical barriers between the VR device and the patient such as combinations of silicon covers, disposable eye masks, wipeable replacement head straps, and hair covers, while 3 (33%) participants did not use barriers (Table 1). Only 1 (11%) institution had an SOP for use and disinfection. Infection prevention teams were involved in assisting with VR disinfection protocols at 2 (22%) sites.\",\n \"A total of 175 experiments were performed to assess disinfection (Table 2). Adequate disinfection was achieved with both the IPA and the alcohol-free quaternary ammonium wipes across all bacterial types and headset material comparing untreated with disinfected surfaces. No bacteria were recovered in 88% (154/175) of experiments. IPA wipes performed better than the quaternary ammonium wipes at reducing overall bacterial counts (P=.001). This difference was most pronounced on porous surfaces, where the mean quantity of bacteria remaining after the alcohol-free quaternary ammonium use was more than after IPA use (Figure 2A).\\nWhen comparing the disinfection by an organism, there were more CFUs of S aureus and S epidermidis recovered after attempted disinfection than P aeruginosa (P=.05 and P=.03, respectively), with this difference most pronounced on porous surfaces (Figure 2B). There was no significant difference between the recovery of S aureus and S epidermidis (P=.72). Finally, fewer bacteria were recovered from porous surfaces after inoculation but prior to performing disinfection. However, more bacteria were recovered after disinfection from porous surfaces than from nonporous surfaces (P=.01) as shown in Figure 2C, confirming that porous materials inoculated (Figure 1) were more difficult to disinfect because of their availability to absorb the bacteria-containing liquid.\\nBacterial count by disinfection method, organism, and surface.\\naThe percentage of times when 0 bacterial colony-forming units were observed.\\nbNonporous surfaces were the Quest 1 Strap, Quest 2 Headset, and controller, while the porous surfaces were the Quest 1 Headset and Quest 2 Strap.\\nThe mean (SEM) raw bacterial counts recovered after disinfection according to disinfectant and equipment (A), by organism and equipment (B), and by equipment surface type (C) are displayed. Note: Non-porous surfaces were the Quest 1 Strap, Quest 2 Headset and Controller, while porous surfaces were the Quest 1 Headset and Quest 2 Strap.\",\n \"Adequate disinfection of VR devices can be achieved through the use of low-level EPA-approved hospital-grade disinfectants commonly used in clinical settings for devices that are exposed to intact skin, including IPA and quaternary ammonium wipes. We found a greater than 6-fold logarithmic reduction from initial bacteria inoculation across all pathogen types and VR device surfaces when using either product. However, we did observe the differences when evaluating raw-bacterial counts after disinfection. Notably, IPA performed better than the quaternary ammonium wipe, particularly for porous surfaces. It is possible that IPA penetrates porous surfaces better than quaternary ammonium products due to the vigorous 15-second scrub, and future studies should evaluate how well different wipes perform on these types of surfaces. We also observed that S aureus and S epidermidis persisted on surfaces at greater densities than P aeruginosa, possibly reflective of a mechanism in strain type or environmental survivability. Finally, and perhaps most critically, we observed lower bacterial counts after inoculation but before disinfection, and greater bacterial counts after disinfection, on porous surfaces when compared to nonporous surfaces. This suggests that bacteria may be entering the pores in the material, potentially reducing exposure to the disinfection material. Additionally, using swabs to recover bacteria from porous surfaces is suboptimal as we do not recover bacteria that have penetrated deeper into the material as well as nonporous surfaces which have better transfer efficiency [10]. Thus, despite consistent recovery after disinfection from porous surfaces in these experiments, we likely have overestimated the efficacy of disinfection for this material. Of note, there was outstanding disinfection of all nonporous surfaces, making it the preferred material for the construction of VR devices in health care. Manufacturers should consider material in the design of both headsets and straps, and our data support the use of nonporous material, particularly in health care settings where persistent bacteria may serve as a nidus for transmission to the next VR device user. If porous surfaces are present, there should be adequate barrier protection to prevent the transmission of microbes. This also then allows for the use of IPA without concern for damage to any porous components of the device.\\nNosocomial transmission and outbreaks associated with the use of medical devices are well documented and a primary concern for using VR devices in health care settings [7]. Consequently, facilities may restrict VR devices from patient use out of concern. We found substantial variability between facilities in the frequency of device use, disinfection method, and barrier protection used. Importantly, almost all sites reported that infection prevention teams were not involved in performing a risk assessment for device use during patient care, and SOPs for disinfection were absent in all but one institution. Establishing a standard process that appropriately disinfects VR devices to allow for safe and expanded use in health care settings while avoiding equipment degradation can benefit patients and health care workers alike. It is critical to ensure that when new devices such as VR equipment are introduced into patient care settings, Infection Prevention and other stakeholders are involved prior to the purchase of the devices to ensure there is an acceptable plan for device reprocessing.\\nBased on this generated data set, manufacturer’s instructions for use [8], health care infection prevention best practices, and previous literature or expert opinion [6], we propose an SOP for VR use and disinfection in the health care setting (Textbox 1). This is particularly important as the patient population served may be undergoing chemotherapy or other immunosuppressants that can increase the risk of infection.\\n\\nBefore use\\n\\nAvoid on patients with nonintact skin or active infections on the head or hands that cannot be covered and might come into contact with the device\\nAvoid use on patients known to be colonized with pathogens where specialized disinfection is required, including Clostridioides difficile, Candida auris, Mycobacterium tuberculosis, and nonenveloped viruses\\nPatient and staff perform hand hygiene\\nA nonporous cover over the face pad, a disposable face pad cover, or both should serve as a barrier between the patient’s face and the device. Hair should also be covered (eg, bouffant and washable cloth surgical cap). Any porous material that makes contact with the patient’s skin or hair should be covered with a barrier\\nDevices should be assessed for alcohol compatibility. If the device is not alcohol compatible, a nonalcohol-based disinfectant should be used\\nPerform disinfection with a device compatible Environmental Protection Agency–registered product List H [11] according to the manufacturer’s instructions for use, ensuring that all surfaces of the headset (including the strap, the casing, the inner and outer facepieces, and the lens), the controller, and the nonporous, nondisposable face cover are saturated. Do not use wipes on multiple devices.\\n\\nAfter use\\n\\nPatient and staff perform hand hygiene\\nStaff don appropriate personal protective equipment, which should include nitrile gloves at a minimum unless other personal protective equipment is required per the patient’s transmission-based isolation precautions\\nRemove the device from the patient and placed on a clean disposable pad\\nDiscard the disposable face cover, if present\\nRemove nonporous, nondisposable face cover from the device, if present\\nClean all visibly soiled areas with disposable wipes or paper towels\\nRepeat disinfection as above\\nAllow headset and controllers to dry according to the product instructions for use\\nStore the device in a dry space physically separated from nondisinfected devices\\nPatient and staff perform hand hygiene\\nIn creating the SOP (Textbox 1), we considered VR devices a noncritical item requiring low-level disinfection between patients because they are most commonly exposed to intact skin only. Given disinfection success with the IPA and alcohol-free quaternary ammonium wipes, we suspect other equivalent low-level disinfectant products (eg, combination IPA or quaternary ammonium wipes) would be adequate, especially when applied to nonporous services [12]. We did not evaluate high-level disinfectants or sterilization procedures that may be required in the event of device exposure to nonintact skin or mucous membranes. We suggest avoiding VR device use on patients who have breaks in the skin on the hands or head region that cannot be appropriately covered and could come into contact with the device, thus avoiding the need for a high level of disinfection.\\nAs disinfection was successfully achieved for a variety of pathogens, VR use is most likely safe for patients where contact isolation (gowns and gloves) is required in the hospital, including patients colonized with methicillin-resistant S aureus or vancomycin-resistant Enterococci. However, we suggest clinicians exercise caution when using VR devices with patients colonized with harder-to-eradicate pathogens such as Clostridioides difficile, Candida auris, and nonenveloped viruses where sodium hypochlorite or other high-level disinfection methods may be required. The pathogens tested (S aureus, P aeruginosa, and S epidermidis) are very common organisms in health care settings seen in both the adult and pediatric populations, and these results from disinfection are likely to be applicable to most health care settings, regardless of the patient’s age.\\nLimitations of this study include the single-site nature limiting generalizability and the poor survey response rate. We also only tested 2 VR devices from the same company and 2 methods of surface disinfection. Preliminary experiments with an ultraviolet C (UVC) device specifically designed to decontaminate VR devices failed to produce adequate disinfection (data not shown). Since UVC disinfection depends on the angle and distance from the surface to the UVC source [13], the geometry of the headsets may make UVC disinfection more challenging. Further studies evaluating alternative disinfection methods, including UVC and other types of VR devices, are ongoing.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":["introduction","methods",null,null,null,null,"results",null,null,"discussion"],"string":"[\n \"introduction\",\n \"methods\",\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n \"discussion\"\n]"},"keywords":{"kind":"list like","value":["disinfection","healthcare-acquired infection","healthcare worker","virtual reality","disinfect","occupational health","occupational safety","infection control","infection spread"],"string":"[\n \"disinfection\",\n \"healthcare-acquired infection\",\n \"healthcare worker\",\n \"virtual reality\",\n \"disinfect\",\n \"occupational health\",\n \"occupational safety\",\n \"infection control\",\n \"infection spread\"\n]"}}},{"rowIdx":388,"cells":{"title":{"kind":"string","value":"The psychological consequences of living with coronary heart disease: Are patients' psychological needs served? A mixed-method study in Germany."},"pmid":{"kind":"string","value":"36269637"},"background_abstract":{"kind":"string","value":"This mixed-method study explores psychological needs, access and barriers in coronary heart disease (CHD) patients with and without mental health issues (MHI) within the German healthcare system."},"background_abstract_label":{"kind":"string","value":"INTRODUCTION"},"methods_abstract":{"kind":"string","value":"This study was conducted in three different healthcare settings: two hospitals, two rehabilitation clinics and three cardiology practices in Cologne, Germany. Patients were screened for angiographically documented CHD and other inclusion criteria. In total, 364 CHD patients took part in this study. It consisted of two parts: In the first part, participants filled in a newly developed questionnaire about their psychological needs, access and barriers within the healthcare system and their contact with their doctor in these matters. Then, patients were screened for MHIs with the help of the Hospital Anxiety and Depression Scale (HADS). When a score above seven was scored on the HADS, patients were additionally screened for specific MHIs using the Structured Clinical Interview for DSM-IV Axis I Disorders. In the second part, 20 participants were subsequently interviewed in a semi-structured interview to generate more in-depth findings."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The interviews show that CHD patients with and without MHI experienced a cardiac event as life-changing and had an urgent need to talk about CHD with their doctor, mostly the general practitioner (GP). When the GP spoke to the patient shortly after the cardiac event, patients experienced relief and were better able to cope with their illness. Only 9.1% reported being aided in their search for psychotherapeutic treatment or drug treatment (4.1%)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"The needs of CHD patients with and without MHI were not adequately satisfied within our sample. Psychological measures are necessary for sufficient improvement, such as training of doctors in doctor-patient communication (e.g., better support in coping with MHI/CHD), improvements in the procedure (more time for conversations during doctor contacts), and improvement of structural requirements (referring patients faster to psychotherapists)."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Coronary Disease","Research Design","Germany","Adaptation, Psychological","Surveys and Questionnaires"],"string":"[\n \"Humans\",\n \"Coronary Disease\",\n \"Research Design\",\n \"Germany\",\n \"Adaptation, Psychological\",\n \"Surveys and Questionnaires\"\n]"},"pmcid":{"kind":"string","value":"9700170"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"RESULTS"},"results_text":{"kind":"string","value":"[SUBTITLE] Sociodemographic and clinical characteristics [SUBSECTION] Most participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.\nMost participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.\n[SUBTITLE] Patients' needs [SUBSECTION] A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.\nA total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.\n[SUBTITLE] Access and barriers to healthcare [SUBSECTION] A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\nA total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\n[SUBTITLE] Qualitative [SUBSECTION] In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\nIn total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\n[SUBTITLE] Perception of the cardiac event and approaches by doctors to talk about MHI [SUBSECTION] The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\nThe following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n[SUBTITLE] Patients' needs [SUBSECTION] [SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144"},"conclusions_title":{"kind":"string","value":"CONCLUSION"},"conclusions_text":{"kind":"string","value":"This study found that the needs regarding CHD patients with and without MHI were not adequately satisfied. Furthermore, for CHD patients with MHI, access and barriers to the healthcare system regarding psychotherapeutic and drug therapy were insufficient. CHD patients experienced the cardiac event as life‐changing and had an urgent need to talk about mental problems. Primarily, the GP was named as a partner for conversations about CHD‐related mental problems; however, when the doctor spoke to the patient shortly after the cardiac event, patients felt relieved and coped better with their illness. To improve this, psychological measures such as training of doctors in doctor–patient communication are necessary, as well as improvements in procedural and structural requirements."},"other_sections_titles":{"kind":"list like","value":["INTRODUCTION","Coronary heart disease","Mental health issues and their relationships with CHD","MHIs as barrier to medical adherence","Current study","Design","Patient contribution","Participants","Research tools","Quantitative","Qualitative","Procedures","Quantitative","Qualitative","Data analysis","Quantitative","Qualitative","Sociodemographic and clinical characteristics","Patients' needs","Access and barriers to healthcare","Qualitative","Perception of the cardiac event and approaches by doctors to talk about MHI","Patients' needs","The wish for a conversation about MHI","No wish for a conversation about MHI","Access and barriers","Specific age‐related issues","The younger patient group and their concerns about their ability to work","The older patient group and their concerns about successful aging","Perception of the cardiac event and patients' needs","Access and barriers to healthcare and specific issues of patient groups","Strengths, limitations and further research","AUTHOR CONTRIBUTIONS"],"string":"[\n \"INTRODUCTION\",\n \"Coronary heart disease\",\n \"Mental health issues and their relationships with CHD\",\n \"MHIs as barrier to medical adherence\",\n \"Current study\",\n \"Design\",\n \"Patient contribution\",\n \"Participants\",\n \"Research tools\",\n \"Quantitative\",\n \"Qualitative\",\n \"Procedures\",\n \"Quantitative\",\n \"Qualitative\",\n \"Data analysis\",\n \"Quantitative\",\n \"Qualitative\",\n \"Sociodemographic and clinical characteristics\",\n \"Patients' needs\",\n \"Access and barriers to healthcare\",\n \"Qualitative\",\n \"Perception of the cardiac event and approaches by doctors to talk about MHI\",\n \"Patients' needs\",\n \"The wish for a conversation about MHI\",\n \"No wish for a conversation about MHI\",\n \"Access and barriers\",\n \"Specific age‐related issues\",\n \"The younger patient group and their concerns about their ability to work\",\n \"The older patient group and their concerns about successful aging\",\n \"Perception of the cardiac event and patients' needs\",\n \"Access and barriers to healthcare and specific issues of patient groups\",\n \"Strengths, limitations and further research\",\n \"AUTHOR CONTRIBUTIONS\"\n]"},"other_sections_texts":{"kind":"list like","value":["[SUBTITLE] Coronary heart disease [SUBSECTION] Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n\nCoronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n\n[SUBTITLE] Mental health issues and their relationships with CHD [SUBSECTION] It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n\nIt has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n\n[SUBTITLE] MHIs as barrier to medical adherence [SUBSECTION] MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n\nMHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n\n[SUBTITLE] Current study [SUBSECTION] In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\nIn the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.","Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n","It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n","MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n","In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.","MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\n16\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\n17\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.","We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.","As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\n18\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\n19\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\n20\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.","[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’","Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n","The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’","[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.","The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.","The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.","[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\nAll presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.","All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.","The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.","Most participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.","A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.","A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.","In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.","The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137","[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144","All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210","A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72","Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96","Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.","The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50","On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144","Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\n11\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\n11\n\n\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\n31\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.","In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\n7\n and hinder the required lifestyle changes to reduce CHD risk factors.\n4\n, \n8\n\n\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\n32\n Following Porter,\n33\n\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\n32\n, \n33\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\n33\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\n34\n\n","MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.","Samia Peltzer wrote the abstract, introduction, method section, results (incl. analyses) and discussion, designed the figure and tables, references with EndNote, made an overall adaption of the text and included references. Samia Peltzer was a major contributor in writing the manuscript. Ursula Köstler conducted the information for the qualitative analysis, conducted qualitative analysis and summed up the results for the qualitative part. Frank Jessen, Frank Schulz‐Nieswandt and Christian Albus designed and conducted the study as chief scientists. Hendrik Müller, Nadine Scholten, Frank Jessen and Christian Albus critically revised the manuscript. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy."],"string":"[\n \"[SUBTITLE] Coronary heart disease [SUBSECTION] Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\\nCoronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\\n[SUBTITLE] Mental health issues and their relationships with CHD [SUBSECTION] It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\\nIt has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\\n[SUBTITLE] MHIs as barrier to medical adherence [SUBSECTION] MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\\nMHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\\n[SUBTITLE] Current study [SUBSECTION] In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\\nIn the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\",\n \"Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\",\n \"It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\",\n \"MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\",\n \"In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\",\n \"MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\\n16\\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\\n17\\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\",\n \"We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\",\n \"As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\\n18\\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\\n19\\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\\n20\\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\",\n \"[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\",\n \"Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\",\n \"The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\",\n \"[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\",\n \"The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\",\n \"The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\",\n \"[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\nAll presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\",\n \"All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\",\n \"The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\",\n \"Most participants were male (n = 258, 70.9%). The mean age across genders was M\\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\\nSociodemographic and clinical characteristics of the participants\\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\\nIt is possible for patients to have more than one answer.\\nAt least one cell was too small for the appropriate analysis.\",\n \"A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\\nAbbreviation: CHD, coronary heart disease.\\nMultiple‐choice question.\\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\\nAt least one cell was too small for the appropriate analysis.\",\n \"A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\",\n \"In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\",\n \"The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\n\\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\",\n \"[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\",\n \"All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\",\n \"A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\",\n \"Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\",\n \"Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\",\n \"The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\",\n \"On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\",\n \"Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\\n11\\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\\n11\\n\\n\\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\\n31\\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\",\n \"In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\\n7\\n and hinder the required lifestyle changes to reduce CHD risk factors.\\n4\\n, \\n8\\n\\n\\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\\n32\\n Following Porter,\\n33\\n\\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\\n32\\n, \\n33\\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\\n33\\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\\n34\\n\\n\",\n \"MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.\",\n \"Samia Peltzer wrote the abstract, introduction, method section, results (incl. analyses) and discussion, designed the figure and tables, references with EndNote, made an overall adaption of the text and included references. Samia Peltzer was a major contributor in writing the manuscript. Ursula Köstler conducted the information for the qualitative analysis, conducted qualitative analysis and summed up the results for the qualitative part. Frank Jessen, Frank Schulz‐Nieswandt and Christian Albus designed and conducted the study as chief scientists. Hendrik Müller, Nadine Scholten, Frank Jessen and Christian Albus critically revised the manuscript. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["INTRODUCTION","Coronary heart disease","Mental health issues and their relationships with CHD","MHIs as barrier to medical adherence","Current study","MATERIALS AND METHODS","Design","Patient contribution","Participants","Research tools","Quantitative","Qualitative","Procedures","Quantitative","Qualitative","Data analysis","Quantitative","Qualitative","RESULTS","Sociodemographic and clinical characteristics","Patients' needs","Access and barriers to healthcare","Qualitative","Perception of the cardiac event and approaches by doctors to talk about MHI","Patients' needs","The wish for a conversation about MHI","No wish for a conversation about MHI","Access and barriers","Specific age‐related issues","The younger patient group and their concerns about their ability to work","The older patient group and their concerns about successful aging","DISCUSSION","Perception of the cardiac event and patients' needs","Access and barriers to healthcare and specific issues of patient groups","Strengths, limitations and further research","CONCLUSION","AUTHOR CONTRIBUTIONS"],"string":"[\n \"INTRODUCTION\",\n \"Coronary heart disease\",\n \"Mental health issues and their relationships with CHD\",\n \"MHIs as barrier to medical adherence\",\n \"Current study\",\n \"MATERIALS AND METHODS\",\n \"Design\",\n \"Patient contribution\",\n \"Participants\",\n \"Research tools\",\n \"Quantitative\",\n \"Qualitative\",\n \"Procedures\",\n \"Quantitative\",\n \"Qualitative\",\n \"Data analysis\",\n \"Quantitative\",\n \"Qualitative\",\n \"RESULTS\",\n \"Sociodemographic and clinical characteristics\",\n \"Patients' needs\",\n \"Access and barriers to healthcare\",\n \"Qualitative\",\n \"Perception of the cardiac event and approaches by doctors to talk about MHI\",\n \"Patients' needs\",\n \"The wish for a conversation about MHI\",\n \"No wish for a conversation about MHI\",\n \"Access and barriers\",\n \"Specific age‐related issues\",\n \"The younger patient group and their concerns about their ability to work\",\n \"The older patient group and their concerns about successful aging\",\n \"DISCUSSION\",\n \"Perception of the cardiac event and patients' needs\",\n \"Access and barriers to healthcare and specific issues of patient groups\",\n \"Strengths, limitations and further research\",\n \"CONCLUSION\",\n \"AUTHOR CONTRIBUTIONS\"\n]"},"all_sections_texts":{"kind":"list like","value":["[SUBTITLE] Coronary heart disease [SUBSECTION] Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n\nCoronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n\n[SUBTITLE] Mental health issues and their relationships with CHD [SUBSECTION] It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n\nIt has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n\n[SUBTITLE] MHIs as barrier to medical adherence [SUBSECTION] MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n\nMHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n\n[SUBTITLE] Current study [SUBSECTION] In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\nIn the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.","Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\n1\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\n2\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\n1\n, \n3\n, \n4\n\n","It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\n5\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\n6\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\n7\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\n8\n\n","MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\n4\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\n9\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\n10\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\n11\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\n12\n\n\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\n4\n, \n13\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\n14\n, \n15\n\n","In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.","This study (MenDis‐CHD) is one of three projects carried out within the framework of the Federal Ministry of Education and Research (BMBF)‐funded Cologne Research and Practice Network (CoRe‐Net). MenDis‐CHD was approved by the Ethics Commission of Cologne University's Faculty of Medicine (Committee Reference Number: 17–220) on 26 September 2017.\n[SUBTITLE] Design [SUBSECTION] MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\n16\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\n17\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\nMenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\n16\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\n17\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\n[SUBTITLE] Patient contribution [SUBSECTION] We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\nWe received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\n[SUBTITLE] Participants [SUBSECTION] As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\n18\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\n19\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\n20\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\nAs inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\n18\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\n19\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\n20\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\n[SUBTITLE] Research tools [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\n[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\n[SUBTITLE] Procedures [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\n[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\n[SUBTITLE] Data analysis [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\nAll presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\n[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\nAll presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.","MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\n16\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\n17\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.","We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.","As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\n18\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\n19\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\n20\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.","[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’","Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\n21\n\n\nThe HADS\n22\n, \n23\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\n24\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\n25\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\n6\n\n","The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\n16\n The qualitative study design was conducted according to the method of the problem‐centred interview.\n26\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’","[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.","The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.","The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.","[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\nAll presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.","All presented data were analysed using IBM SPSS Statistics 22 software\n27\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.","The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\n28\n for content analysis.\n29\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.","[SUBTITLE] Sociodemographic and clinical characteristics [SUBSECTION] Most participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.\nMost participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.\n[SUBTITLE] Patients' needs [SUBSECTION] A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.\nA total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.\n[SUBTITLE] Access and barriers to healthcare [SUBSECTION] A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\nA total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\n[SUBTITLE] Qualitative [SUBSECTION] In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\nIn total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\n[SUBTITLE] Perception of the cardiac event and approaches by doctors to talk about MHI [SUBSECTION] The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\nThe following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n[SUBTITLE] Patients' needs [SUBSECTION] [SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144","Most participants were male (n = 258, 70.9%). The mean age across genders was M\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\nSociodemographic and clinical characteristics of the participants\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\nIt is possible for patients to have more than one answer.\nAt least one cell was too small for the appropriate analysis.","A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\nAbbreviation: CHD, coronary heart disease.\nMultiple‐choice question.\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\nAt least one cell was too small for the appropriate analysis.","A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.","In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.","The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\n\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137","[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144","All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\n\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210","A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\n\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72","Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\n\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96","Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.","The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\n\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50","On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\n\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144","This mixed‐method study aimed at exploring needs, access and barriers in the care of CHD patients from the patient's viewpoint. In our study, 28% of CHD patients were identified as having MHI. In accordance with the literature, the MHI in our study were mainly depression and anxiety disorders,\n7\n, \n8\n, \n30\n and women were twice as likely to develop MHI.\n1\n Taking the quantitative and qualitative results together, patients reported unmet needs, decreased access to helpful resources and encountered multiple obstacles in handling their MHI adequately.\n[SUBTITLE] Perception of the cardiac event and patients' needs [SUBSECTION] Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\n11\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\n11\n\n\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\n31\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\nBoth quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\n11\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\n11\n\n\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\n31\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\n[SUBTITLE] Access and barriers to healthcare and specific issues of patient groups [SUBSECTION] In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\n7\n and hinder the required lifestyle changes to reduce CHD risk factors.\n4\n, \n8\n\n\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\n32\n Following Porter,\n33\n\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\n32\n, \n33\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\n33\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\n34\n\n\nIn general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\n7\n and hinder the required lifestyle changes to reduce CHD risk factors.\n4\n, \n8\n\n\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\n32\n Following Porter,\n33\n\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\n32\n, \n33\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\n33\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\n34\n\n\n[SUBTITLE] Strengths, limitations and further research [SUBSECTION] MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.\nMenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.","Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\n11\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\n11\n\n\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\n31\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.","In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\n7\n and hinder the required lifestyle changes to reduce CHD risk factors.\n4\n, \n8\n\n\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\n32\n Following Porter,\n33\n\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\n32\n, \n33\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\n33\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\n34\n\n","MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.","This study found that the needs regarding CHD patients with and without MHI were not adequately satisfied. Furthermore, for CHD patients with MHI, access and barriers to the healthcare system regarding psychotherapeutic and drug therapy were insufficient. CHD patients experienced the cardiac event as life‐changing and had an urgent need to talk about mental problems. Primarily, the GP was named as a partner for conversations about CHD‐related mental problems; however, when the doctor spoke to the patient shortly after the cardiac event, patients felt relieved and coped better with their illness. To improve this, psychological measures such as training of doctors in doctor–patient communication are necessary, as well as improvements in procedural and structural requirements.","Samia Peltzer wrote the abstract, introduction, method section, results (incl. analyses) and discussion, designed the figure and tables, references with EndNote, made an overall adaption of the text and included references. Samia Peltzer was a major contributor in writing the manuscript. Ursula Köstler conducted the information for the qualitative analysis, conducted qualitative analysis and summed up the results for the qualitative part. Frank Jessen, Frank Schulz‐Nieswandt and Christian Albus designed and conducted the study as chief scientists. Hendrik Müller, Nadine Scholten, Frank Jessen and Christian Albus critically revised the manuscript. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy."],"string":"[\n \"[SUBTITLE] Coronary heart disease [SUBSECTION] Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\\nCoronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\\n[SUBTITLE] Mental health issues and their relationships with CHD [SUBSECTION] It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\\nIt has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\\n[SUBTITLE] MHIs as barrier to medical adherence [SUBSECTION] MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\\nMHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\\n[SUBTITLE] Current study [SUBSECTION] In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\\nIn the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\",\n \"Coronary heart disease (CHD) is the leading cause of mortality, accounting for approximately one‐third of all deaths in individuals over the age of 35 years, with a death rate of 102.6 per 100,000.\\n1\\n In Germany, the lifetime prevalence of CHD for the age group 40−79 years is 9.3% (8.4−10.3).\\n2\\n Short‐term mortality has been decreased due to advances in acute CHD treatments. However, at a population level, CHD morbidity has increased in recent years. After an interim decline in smoking, hypercholesterolaemia and hypertension, these currently appear to be on the rise again.\\n1\\n, \\n3\\n, \\n4\\n\\n\",\n \"It has been found that mental health issues (MHIs) are highly prevalent among patients with CHD. MHI is defined as a recognizable set of clinical symptoms and/or behavioural problems that correlate with individual distress and impairment of functioning at the individual level.\\n5\\n After a myocardial infarction, almost 30% of these patients suffer from depressive symptoms and 20% even fulfil ICD‐10 criteria for a depressive episode.\\n6\\n Depression is associated with higher CHD‐related morbidity and mortality (relative risk = 1.6–2.4): After a cardiac event such as a myocardial infarction, CHD patients with comorbid depression have twofold greater mortality during the next 2 years.\\n7\\n Furthermore, mortality risk can increase sixfold when the patient suffers from severe depressive symptoms.\\n8\\n\\n\",\n \"MHI also acts as a strong barrier to treatment adherence and exacerbates a required lifestyle change that is necessary to reduce CHD risk factors—regardless of whether MHI were already pre‐existing or occurred as a consequence of CHD.\\n4\\n For example, patients with MHI are less likely to exhibit healthy lifestyle behaviour (e.g., quit smoking) and have low medication adherence.\\n9\\n This emphasizes the drive for more patient involvement in terms of disease management and medical consultations.\\n10\\n At the same time, this group of patients finds it difficult to understand the course of the consultation and to present their own concerns adequately, resulting in a lack of communication of possible physical and/or MHIs to their doctor.\\n11\\n Hence, this specific population of CHD patients with MHI may be very vulnerable due to difficulties in disease management, low communication skills in a medical setting and problems communicating possible issues and questions about personal topics. This suggests that the quality of communication between patients and their doctors is important to secure the best course of treatment.\\n12\\n\\n\\nConsequently, adequate screening and treatment for MHI in CHD patients during the medical consultation from the doctor's perspective is urgently needed. National and international guidelines have recommended routine screening and treatment for MHI on primary and secondary CHD prevention.\\n4\\n, \\n13\\n However, studies have been able to show that this screening is not regularly followed due to lack of time and low occurrence of verbal interventions.\\n14\\n, \\n15\\n\\n\",\n \"In the current study, we explore the following research questions: (1) What are the MHI‐related needs of CHD patients with MHI after a cardiac event?, (2) how well do doctors meet their patients' needs? and (3) how do patients experience access and barriers to the healthcare system? It is hypothesized that patients with MHI have a desire for doctors to approach and support them with their MHI. Secondly, it is expected that doctors rarely meet their patients' needs. Lastly, it is hypothesized that, due to Hypotheses 1 and 2, patients receive a lack of access and distinct barriers to the healthcare system, especially related to their MHI. The aim of the study is to improve the quality of care and initiate changes in the structure of healthcare towards more patient‐centred care in the long term.\",\n \"This study (MenDis‐CHD) is one of three projects carried out within the framework of the Federal Ministry of Education and Research (BMBF)‐funded Cologne Research and Practice Network (CoRe‐Net). MenDis‐CHD was approved by the Ethics Commission of Cologne University's Faculty of Medicine (Committee Reference Number: 17–220) on 26 September 2017.\\n[SUBTITLE] Design [SUBSECTION] MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\\n16\\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\\n17\\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\\nMenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\\n16\\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\\n17\\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\\n[SUBTITLE] Patient contribution [SUBSECTION] We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\\nWe received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\\n[SUBTITLE] Participants [SUBSECTION] As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\\n18\\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\\n19\\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\\n20\\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\\nAs inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\\n18\\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\\n19\\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\\n20\\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\\n[SUBTITLE] Research tools [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\n[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\n[SUBTITLE] Procedures [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\n[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\n[SUBTITLE] Data analysis [SUBSECTION] [SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\nAll presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\\n[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\nAll presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\",\n \"MenDis‐CHD is a cross‐sectional study with a mixed‐method approach using an explanatory sequential design,\\n16\\n beginning with quantitative data collection (one questionnaire containing: questions about sociodemographic data, patients' needs and access and barriers to the healthcare system; psychological assessments [Hospital Anxiety and Depression Scale {HADS}, SCID‐I]), followed by a qualitative data acquisition (semi‐structured in‐depth interviews) to enrich the quantitative data. Specific details of the methods used in this study can be found in an earlier paper.\\n17\\n The questionnaire was fully developed in German. Individual translated items can be provided upon request.\",\n \"We received input from CHD patients during the pretests of the questionnaire and the qualitative interviews, as well as from members of CHD self‐help groups. The patient perspective has been continuously taken into account in the project and the interests of the patients have also been considered with regard to the transfer of results (e.g., through public lectures, summaries in plain language and design of a leaflet to share with peers and distribute to patient groups). Patients did not receive any compensation for contributing to this study.\",\n \"As inclusion criteria, patients had to have an angiographically documented CHD with stable angina pectoris, acute coronary syndromes, percutaneous coronary intervention (PCI) or bypass surgery, be older than 18 years, have sufficient German language skills and be able to give informed consent. Severe physical (e.g., cancer) or unstable mental problems (e.g., acute suicidal ideation, delirium and moderate to severe dementia) were the exclusion criteria.\\nOverall, 753 patients were screened for eligibility, of whom 374 were finally enroled in the study. Ten patients dropped out due to withdrawal of informed consent or incomplete questionnaires with missing answers. Therefore, 364 patients were included in the final data analysis. We recruited 107 patients in hospitals, 157 patients in rehabilitation clinics and 100 patients in cardiology practices. Figure 1 features a flowchart presenting the recruitment process.\\nFlowchart of the recruitment procedure. CHD, coronary heart disease; LVEF, left ventricular ejection fraction; MHI, mental health issue\\nAll 364 included patients filled in the questionnaire. We used maximum variation as the sampling strategy to increase diversity by age, gender, marital status, employed/retired status, severity of CHD, left ventricular ejection fraction and kind and severity of diagnosis of MHI.\\n18\\n By explicitly aiming for recruiting a diverse sample, one ensures the generalizability of one's findings across a broader part of the general population. Furthermore, we asked every eligible and interested patient during the recruitment phase to also participate in our in‐depth interviews. In total, 20 patients were selected to participate in the qualitative interviews. The diversity of participants was actively monitored, but due to a significant number of patients who did not want to participate in the interviews, participant diversity was not completely balanced, especially gender and ethnicity. In this study, the sample consisted of approximately 70% males. According to the current literature, males are significantly more likely to develop CHD,\\n19\\n with a lifetime prevalence of 9.9% (9.0–10.8) than women, with 6.7% (6.0–7.4).\\n20\\n Thus, this study seems to represent an appropriate sample population. For further details about the distribution of samplings for interview groups, see Figure 1.\",\n \"[SUBTITLE] Quantitative [SUBSECTION] Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\nSociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\\n[SUBTITLE] Qualitative [SUBSECTION] The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\\nThe qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\",\n \"Sociodemographic data were assessed via a newly developed questionnaire. Disease severity of CHD patients was assessed by screening for cardiac events in previous medical history (e.g., myocardial infarction), cardiac surgeries (e.g., bypass surgery), congestive heart failure, somatic comorbidity, pre‐existing presence of MHI, left ventricular ejection fraction and New York Heart Association stage (classification scheme of heart diseases according to their degree of severity). All of these characteristics were drawn from patients' medical records; they were not part of the questionnaire. Clinical data were measured with the use of the HADS and SCID‐I.\\nAn aforementioned self‐made questionnaire was made of five subsections of which two are used in this paper, namely ‘Patients' needs’ and ‘Access and barriers to healthcare system’. The questionnaire was developed and used in the German language only; all items and section headers provided in this article are translated.\\nBoth sections, ‘Patients' needs’ and ‘Access and barriers to the healthcare system,’ were assessed by a 20‐item self‐report questionnaire. Patients without MHI could skip several items that contained questions about an existing MHI. The section ‘Patient's needs’ contained six questions, such as ‘Do you wish to be approached by the doctor on possible mental health issues?’ with binary answer possibilities (Yes/No) or multiple‐choice questions such as ‘Who do you think should give a recommendation for treatment in the case of mental health issues?’. The section ‘Access and barriers to healthcare system’ contained 14 questions, also designed with ‘yes or no’ or multiple‐choice questions such as ‘How often have you sufficiently received information on the content and accessibility of psychological care?’. The questionnaire was developed within this study with the help of four CHD patients with and without MHI and has yet to be validated. More information on the content of the questionnaire and how it was created can be found in our first publication on MHI.\\n21\\n\\n\\nThe HADS\\n22\\n, \\n23\\n was used as a general screening tool to assess symptoms of depression and anxiety and was filled in by the patients. In line with the advice of several studies,\\n24\\n a score of 8 and above was considered a ‘positive’ test result for anxiety or depression. In this case, the Structured Clinical Interview for DSM‐IV Axis I Disorders (SCID‐I)\\n25\\n was conducted as a semi‐structured interview to assess mental symptoms and disorders as defined in the ICD‐10.\\n6\\n\\n\",\n \"The qualitative module was designed to build on the quantitative part in a second phase to enrich the quantitative findings and further investigate the research questions of the study.\\n16\\n The qualitative study design was conducted according to the method of the problem‐centred interview.\\n26\\n This is a semi‐structured in‐depth interview approach that allows the respondent to speak as freely as possible but is centred on (a) predefined problem(s) to which the interviewer repeatedly refers back during the conversation. Based on the research questions of this study, three main categories were developed: (1) Category ‘Perception of the cardiac event and approaches of doctors to talk about MHI’: Were CHD patients approached by the doctor on MHI and after a cardiac event? How did patients experience this approach? What were their expectations? What were positive and negative aspects of the contact with the doctor? Example questions were ‘Have you been approached by the doctor after a cardiac event and if so, how did you experience it? How long did it take until you were asked? Which aspects have been discussed during your meeting? What were your expectations beforehand and what are your expectations now?’; (2) Category ‘Patient's needs’: Do patients want to talk about MHI? Who do patients want to talk to about MHI? If a doctor is the first choice, what type of doctor? Example questions: ‘Who asked you about MHI? Who would you have liked to be addressed by? What did you need (both physically and mentally) after you received your diagnosis?’; (3) Category ‘Access and barriers’: How exactly do patients experience access and barriers to the health care system? Example questions: ‘What information and offers of help were accessible? How well could then be implemented? What information or offers did you receive, which did you not receive albeit your wish to receive them?’\",\n \"[SUBTITLE] Quantitative [SUBSECTION] The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\nThe questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\\n[SUBTITLE] Qualitative [SUBSECTION] The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\\nThe interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\",\n \"The questionnaire was pre‐tested on two persons with CHD, one with MHI and one without MHI, in July 2017. Participants were recruited between 15 January 2018 and 29 March 2019 in two hospital cardiac departments, three cardiology practices and two rehabilitation clinics in Cologne, Germany, to achieve an appropriate realistic sample of routine healthcare settings. After screening for eligibility through the researchers, all patients' demographic data were documented in our screening log. Those who fulfilled our inclusion criteria and provided written informed consent received the quantitative questionnaires. A second appointment was arranged to perform the SCID‐I if the HADS screening was ‘positive’. All researchers of this study were trained in applying the SCID‐I and experienced in conducting it. In addition, patients were asked to participate in qualitative interviews.\",\n \"The interview guideline was pretested by two CHD patients: one with MHI and the other without. Pretests were conducted as face‐to‐face interviews in September 2018. Out of 65 participants who agreed to take part in a qualitative interview, 20 were randomly selected. The actual 20 interviews were carried out as face‐to‐face interviews between October 2018 and March 2019, either during a home visit or at the setting where the respective patient was recruited. Informed consent was obtained during the quantitative phase, but the interviewer informed all patients again by reading the information sheet and informed consent to them to ensure that they fully understood each element. The interview took 1 h. All interviews were audio‐recorded and transcribed.\",\n \"[SUBTITLE] Quantitative [SUBSECTION] All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\nAll presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\\n[SUBTITLE] Qualitative [SUBSECTION] The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\\nThe accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\",\n \"All presented data were analysed using IBM SPSS Statistics 22 software\\n27\\n and controlled for both missing and impossible values. Identified cases with either missing or implausible values were compared to their equivalent paper versions and corrected. Multivariate analyses of variance and χ\\n2 tests were conducted, depending on the curvature of the distribution of the variables' scores.\",\n \"The accuracy of the transcripts was checked based on the audiotape samples by two of the authors (a social scientist and a clinical psychologist). Data were imported to F4 Analysis\\n28\\n for content analysis.\\n29\\n In F4 Analysis, the audiotapes of the interviews were automatically converted into text. Then, both researchers listened to the tapes and checked the transcripts independently. Relevant clusters (content‐wise) were developed and presented to the steering group of MenDis‐CHD for interpretation. Data and coding within the different clusters were further analysed, named and defined as: (1) perception of cardiac event; (2) patient's needs; (3) access and barriers to the healthcare system and (4) specific age‐related issues. These categories reflect topics most important to patients during the semi‐structured interview. They were identified through discussion until an agreement was reached. Besides creating an overview of relevant content per category, this procedure was chosen to verify that patients would indeed talk about the research questions, as it was the aim to gather more qualitative information on them. Regular in‐depth discussion meetings of the steering group of MenDis‐CHD increased the reliability of the research results.\",\n \"[SUBTITLE] Sociodemographic and clinical characteristics [SUBSECTION] Most participants were male (n = 258, 70.9%). The mean age across genders was M\\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\\nSociodemographic and clinical characteristics of the participants\\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\\nIt is possible for patients to have more than one answer.\\nAt least one cell was too small for the appropriate analysis.\\nMost participants were male (n = 258, 70.9%). The mean age across genders was M\\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\\nSociodemographic and clinical characteristics of the participants\\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\\nIt is possible for patients to have more than one answer.\\nAt least one cell was too small for the appropriate analysis.\\n[SUBTITLE] Patients' needs [SUBSECTION] A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\\nAbbreviation: CHD, coronary heart disease.\\nMultiple‐choice question.\\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\\nAt least one cell was too small for the appropriate analysis.\\nA total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\\nAbbreviation: CHD, coronary heart disease.\\nMultiple‐choice question.\\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\\nAt least one cell was too small for the appropriate analysis.\\n[SUBTITLE] Access and barriers to healthcare [SUBSECTION] A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\\nA total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\\n[SUBTITLE] Qualitative [SUBSECTION] In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\\nIn total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\\n[SUBTITLE] Perception of the cardiac event and approaches by doctors to talk about MHI [SUBSECTION] The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\n\\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\nThe following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\n\\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\n[SUBTITLE] Patients' needs [SUBSECTION] [SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\",\n \"Most participants were male (n = 258, 70.9%). The mean age across genders was M\\nage = 65.9 (SD = 11.4); 72.5% lived together with a partner (n = 264); 27.5% had a left ventricular ejection fraction of <40%; 63.2% had a PCI for a myocardial infarction; 52.5% (n = 191) received a positive HADS score and 28.0% (n = 102) had a positive SCID diagnosis. Twelve patients had a positive HADS score but refused further testing (SCID‐I). The following SCID diagnoses were found: unipolar depression (n = 59, 16.5%), bipolar disorder (n = 2, 0.6%), anxiety disorder (n = 26, 7.2%), substance use/addiction disorder (n = 14, 4.0%), adjustment disorder (n = 11, 3.0%) and posttraumatic stress disorder (n = 4, 1.1%). Table 1 provides an overview of the demographic and clinical characteristics. For more details on the ICD‐10 diagnoses, see Table A1.\\nSociodemographic and clinical characteristics of the participants\\nAbbreviations: MHI: mental health issue; NYHA; New York Heart Association; PCI, percutaneous coronary intervention.\\nIt is possible for patients to have more than one answer.\\nAt least one cell was too small for the appropriate analysis.\",\n \"A total of 80.8% with MHI and 74.2% without MHI found it appropriate to have been approached by the doctor (e.g., GP or cardiologist) about MHI; several patients were unsure whether they found it appropriate (5.8% with MHI; 7.5% without MHI). Patients expressed the general wish to be approached proactively by their doctor about MHI (46.5% with MHI; 38.0% without MHI). Approximately a fifth of patients in both groups were unsure about whether they wished to be approached by the doctor about MHI. About 23.4% of MHI patients mentioned a preference for GPs to help them find adequate treatment for drug therapy. For detailed information, see Table 2.\\nDetails to access, barriers and needs regarding healthcare in CHD patients with and without mental health issues (MHI)\\nAbbreviation: CHD, coronary heart disease.\\nMultiple‐choice question.\\nAll percentages relate to the maximum number of patients who were eligible for the underlying question.\\nAt least one cell was too small for the appropriate analysis.\",\n \"A total of 30 CHD patients with MHI (42.3%) stated that they received an adequate amount of information on the content and accessibility of psychological care. Only a minority of patients obtained information about psychotherapy (7.0%) with a quarter receiving no information at all. About 30% of patients obtained information on the content and accessibility of psychological care through their doctors, while 35% stated that they did not receive further information. When patients did obtain information, half of all patients with MHI got information from the GP or other persons outside of the healthcare system (e.g., family members, friends); 47.8% of MHI patients and 44.1% of patients without MHI wished to get this specific information from the GP. For further information, see Table 2.\",\n \"In total, 65 CHD patients agreed to participate in a qualitative interview. These patients were mostly men (n = 53) and had an ejection fraction of <40% (n = 41). Divided over the three settings, 21.5% of patients came from hospitals, 33.9% from rehabilitation clinics and 44.6% from cardiology practices. From this subpopulation, 20 patients were selected and asked to participate. The mean age was M = 67.2 (SD = 12.6). All patients lived with a partner. For detailed information about the selection process, see Figure 1.\",\n \"The following section contains information about how the patients experienced the cardiac event and whether they were really approached by the doctor to talk about MHI. Patients experienced CHD as a decisive turning point that unsettled them permanently. Most respondents felt significant anxiety and uncertainty in the weeks following the cardiac event. After diagnosis, spouses and family were the central persons to whom patients articulated their fears. Only a few patients reported that they were approached about MHI by a doctor as part of the CHD diagnosis. In general, MHI was not discussed until months after the diagnosis. Few patients engaged in a conversation about MHI with a professional. They reported that either the GP or a psychologist in a rehabilitation clinic approached them.No, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24Of course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85Well, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\\n\\nNo, not really. No one actually asked, no doctor, whether there was still a need for treatment. Patient 16, l. 22–24\\nOf course, the doctor should have to ask more background information. How the patient is doing in other areas after this disease, whether he has psychological issues. I think that most specialists, e.g., cardiologists, proceed very technocratically and reel off their plan and don't care much about other things. I have not met any cardiologist, except for the head physician at the university hospital, who has dealt more intensively with this issue. For them it is only important how the echography is, how the ultrasound examination is, the blood values, etc. That's it. I even believe that cardiologists don't have a heart (laughs). Patient 9, l. 78–85\\nWell, I have the feeling that my cardiologist doesn't get involved with me on an emotional level. It's relatively quick, he does a ECG or an ultrasound or something else and then I get my information and have to leave. It's not something where you have a longer conversation, that's what they do. only the treatment, then you go back to the treating GP. Patient 14, l. 131–137\",\n \"[SUBTITLE] The wish for a conversation about MHI [SUBSECTION] All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\nAll patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n[SUBTITLE] No wish for a conversation about MHI [SUBSECTION] A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\nA minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n[SUBTITLE] Access and barriers [SUBSECTION] Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\nLimited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n[SUBTITLE] Specific age‐related issues [SUBSECTION] Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\nBesides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\\n[SUBTITLE] The younger patient group and their concerns about their ability to work [SUBSECTION] The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\nThe needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n[SUBTITLE] The older patient group and their concerns about successful aging [SUBSECTION] On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\nOn the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\",\n \"All patients reported that they would find it appropriate to be asked by a doctor about MHI and that they wish for a conversation about MHI with their doctors. Patients reported a high level of suffering until they talked about MHI. If patients had a professional talk about MHI, the important counterpart was the GP. The conversation was experienced as very positive and relieving when GPs talked about psychological problems on their initiative. Patients reported that the professional talk about MHI should occur shortly after the CHD diagnosis. Patients deliberately changed the GP to find a doctor who was willing to talk about MHI or even actively approached them about possible MHI. One patient experienced contact with his cardiologist as very technical and rational and switched to a cardiologist where he also experienced emotional care. Few patients remembered that they were approached for MHI during their stay in rehabilitation clinics. Those who had psychological appointments in rehabilitation clinics experienced them as positive. No further psychological treatment was needed in the subsequent treatment. Interventions had positive effects on patients' coping skills and self‐care but also resulted in feelings of relief.I think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70It has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222But then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\\n\\nI think that all this belongs to a good doctor. Of course, he should carefully address everything. But he should ask. Everything, all these things, so that a depression can be recognized at an early stage. […] In the course of the conversation the doctor might get the impression that there is a serious problem and I think a good doctor should also address that. Patient 8, l. 68–70\\nIt has to be a doctor, a GP, who talks to people. So, we've had a new family doctor for almost two years now, and he talks to me. The previous one didn't talk to me. And I'm very happy that I can go there, where I can just sit for a moment and think about what I still have to talk about, and as far as I know that's not usual. So otherwise, I came in, five minutes,—bang bang—went out of the room, got my prescription and that's it. I just don't want to have things like that anymore. Patient 14, l. 215–222\\nBut then there was a doctor in rehab, that was the first one who really listened to me, and he also got my story, really informed himself, he was the one who told me that everything was connected. The depression, the heart, and so on. This man has helped me very much. He acknowledged that. that this was not overreacting. I was really suffering. Patient 8, l. 206–210\",\n \"A minority of patients did not want to talk about possible psychological distress and denied it. They reported strictly separate physical and psychological impairments or even denied an interaction between the two. One patient denied a link between physiological events such as CHD and possible psychological consequences.No. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\\n\\nNo. For me, everything just continues as before. The heart attack was basically like a car repair. There was a doctor who said I had a constriction in the arteries of the heart where this stent was inserted. And after that, that was it for me, I hardly gave it a second thought. Patient 5, l. 68–72\",\n \"Limited resources (e.g., lack of time, staff shortage) were often reported as barriers. Several interviewed patients reported that during a triweekly stay in a rehabilitation clinic, only a few patients were screened and asked about MHI. Patients with MHI reported numerous unsuccessful attempts to have a psychological appointment and failures to find an ambulant psychotherapist. They complained about insufficient support by the healthcare system, lacking opportunities to talk about their MHI on medical appointments and lacking help with the search for an ambulant psychotherapist. Drug therapy was not a sufficient alternative for this group; most of them rejected this alternative. Patients were demoralized by those failures, hence they developed a passive coping style, stopped actively searching for psychotherapeutic support or postponed it for the future.The rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200Well, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\\n\\nThe rehabilitation was initiated in the hospital and in the rehab itself there also was a psychologist who was supposed to take care of the processing of these things, but unfortunately she was ill while I was in rehab. And then I was not able to implement this in the rehab. I would have liked to have some appointments with her. Now I had bad luck that the psychological support was cancelled. Maybe that would have changed something. Patient 9, 22–26, l. 197–200\\nWell, a drug therapy. I don't believe that this is necessary. Maybe in one case for sure, but I don't think that would help me. Patient 16, l. 94–96\",\n \"Besides the three main categories, a fourth important category emerged and was summed up as ‘specific issues of younger and older patient groups’.\",\n \"The needs of patients varied according to their life phases. Patients in the age range 50–60 years expressed fears and concerns about their future ability to continue with their work. They were also the subgroup who experienced the most depression and anxiety symptoms and reported the highest level of suffering. Employed patients were more accessible to talk about MHI in general, were more sensitive to handle the situation adequately, showed a high adherence to medical advice, were more informed about possible psychotherapeutic services, were more motivated to be active in their medical treatment and had better problem‐solving mechanisms.Well, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51I don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\\n\\nWell, I'm working again now, aren't I? But when it came to the question of when I would go back to work, I was very unhappy and very uncertain and I thought, how can I do that? Because it is … well, it is still that I am anxious. So, I have body sensations that are strange to me. And they frighten me. And then I think, well, I hope I don't fall down and die. Well, that's something like that. And once I wasn't feeling well and my GP said to me: ‘You can live up to a 100 years with the bypass, but if you don't work on yourself or something like that, then it's not good for your heart and you'll get sicker.’ […] I'm concerned about the disabled person's pass. Maybe I should go to a psychiatrist to discuss it. Patient 14, l. 37–51\\nI don't know if you can imagine that you would be told overnight that you can no longer participate in the active working life that you had and could have until then. Patient 7, l. 48–50\",\n \"On the other side, patients who were already retired (65 years and above) expressed more worries about successful ageing. This type of patient often reported fears about the future quality of life and the responsibility for their own family. Retired patients suffered from more somatic comorbidities and were less active in their own medical treatment. For such patients, social networks functioned as a resistance factor. Other factors were being optimistic, acceptance of the current situation and appreciating the last years of life.‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\\n\\n‘I think the family must be informed, especially the closer family circle. Just because I had two heart attacks and I hear badly on the left side. that is a handicap in life that you don't see. […] these are limitations that no one can imagine […] this handling of heart diseases in society … if you are no longer so efficient and so resilient, mentally as well as physically, it really is an impairment. One does not see the disease and it is no problem for an outsider who has no problems with it … (Those people) think to themselves: why can't this man walk five stairs? […] And these are such things that are important for a family, because one must also manage everyday life and have the support to get along again in social life. Patient 16, l. 117–125 and l. 135–144\",\n \"This mixed‐method study aimed at exploring needs, access and barriers in the care of CHD patients from the patient's viewpoint. In our study, 28% of CHD patients were identified as having MHI. In accordance with the literature, the MHI in our study were mainly depression and anxiety disorders,\\n7\\n, \\n8\\n, \\n30\\n and women were twice as likely to develop MHI.\\n1\\n Taking the quantitative and qualitative results together, patients reported unmet needs, decreased access to helpful resources and encountered multiple obstacles in handling their MHI adequately.\\n[SUBTITLE] Perception of the cardiac event and patients' needs [SUBSECTION] Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\\n11\\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\\n11\\n\\n\\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\\n31\\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\\nBoth quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\\n11\\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\\n11\\n\\n\\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\\n31\\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\\n[SUBTITLE] Access and barriers to healthcare and specific issues of patient groups [SUBSECTION] In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\\n7\\n and hinder the required lifestyle changes to reduce CHD risk factors.\\n4\\n, \\n8\\n\\n\\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\\n32\\n Following Porter,\\n33\\n\\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\\n32\\n, \\n33\\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\\n33\\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\\n34\\n\\n\\nIn general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\\n7\\n and hinder the required lifestyle changes to reduce CHD risk factors.\\n4\\n, \\n8\\n\\n\\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\\n32\\n Following Porter,\\n33\\n\\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\\n32\\n, \\n33\\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\\n33\\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\\n34\\n\\n\\n[SUBTITLE] Strengths, limitations and further research [SUBSECTION] MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.\\nMenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.\",\n \"Both quantitative and qualitative data show in general that for all recruitment centres (hospital cardiac departments, cardiology practices and rehabilitation clinics) patients with and without MHI experience anxiety and uncertainty after a cardiac event. This event was perceived as drastic and life‐changing, and consequently, patients felt a great need to talk about what had happened. They expressed disappointment about the fact that doctors rarely address MHI. Interviewed patients stated that they show little initiative during their medical consultation and usually did not address MHI on their own. A study from Lussier et al.\\n11\\n developed an online tool to encourage patients to be more active and prepared for their healthcare encounters by providing skills coaching via a website. Although patients reported a positive perception of such coaching and a higher motivational level, barriers such as lack of interest, limited access to technology, lack of time or language skills reduced the success of this approach.\\n11\\n\\n\\nHowever, when MHI was discussed with the doctor, patients did report a decrease in symptoms, mainly a reduction in anxiety, and a positive influence on their feeling of being able to cope with the cardiac event. It therefore seems reasonable that doctors, for their part, ask more regularly about MHI during medical consultations to improve the previous deficit in the area of emotional care. For this purpose, it seems necessary to have targeted interventions, such as creating greater awareness for this topic and the vulnerable patient group in everyday medical practice, to ensure rapid detection of MHI and thus adequate care. Further training courses, in the field of basic psychosomatic care and medical interviewing, could be helpful to increase individual communication skills in doctor–patient conversations, especially in an empathic manner during medical conversations. Studies showed that empathy of the doctor perceived by the patient significantly improved patient satisfaction with the received medical care, especially in terms of information exchange, treatment, perceived expertize of the doctor, as well as interpersonal trust and partnership with the doctor.\\n31\\n A higher empathy‐related behaviour of the doctor might be useful to improve the doctor–patient relationship and might increase the chance of patients daring to speak about MHI to their doctor.\",\n \"In general, patients received little support in seeking psychological, psychiatric or psychotherapeutic help. Lack of time and personal opportunities, none to little help in coping with mental stress, and finding an ambulant psychotherapist were found to be barriers for patients and GPs in dealing with their CHD‐related MHI's.\\nShortcomings in the early stages of CHD and MHI‐related treatment can lead to demoralization and learned helplessness in patients. At the same time, complaints such as future‐ and health‐related worries, anxiety, depression and fear of incapacity for work continue to exist and further restrict patients' mental and physical health. As mentioned before, MHI and even mental stress are associated with higher CHD‐related morbidity and mortality\\n7\\n and hinder the required lifestyle changes to reduce CHD risk factors.\\n4\\n, \\n8\\n\\n\\nThese shortcomings in medical healthcare can occur because the current healthcare system defines ‘valuable care’ as health outcomes relative to costs and hence encompasses efficiency. However, cost reduction without regard to the achieved outcomes leads to limited effective healthcare.\\n32\\n Following Porter,\\n33\\n\\nvalue‐based healthcare has the following definition: value equals health outcomes that matter to patients regarding the costs of delivering the outcomes. Thus, value measures all services and activities that determine success by meeting the patients' needs. These needs are determined by the medical condition of the patient and are an interrelated set of medical circumstances. A medical condition includes the most commonly associated conditions—meaning that, for example, care for CHD patients must consist of care for hypertension, and so forth.\\n32\\n, \\n33\\n Regarding the high prevalence of MHI in CHD patients, it might be advisable to include MHI as further interrelated medical circumstances. For example, only CHD patients in rehabilitation clinics reported having access to psychotherapeutic help in general; patients from practices and hospitals lacked this possibility. Rehabilitation clinics seem to work more within a more value‐based healthcare cycle because they are medical units that specialize in several diseases and thus are organized in the same way as an integrated practice unit,\\n33\\n thus including all the necessary skilled medical practitioners needed for medical conditions and their co‐occurring problems.\\nTo deal with the shortcomings found in this study, a more value‐based and patient‐centred approach is necessary. With regard to structural requirements, earlier detection and correct diagnosis of MHI, as well as appropriate and fast treatment, would be examples (i.e., an appointment with an outpatient psychotherapist). In addition, more psychological measures, such as training for doctors (e.g., to increase the level of empathy and better support in coping with CHD and MHI) and improvements in procedural requirements (more conversation time during doctor contacts and more precise agreements between different departments) are needed. To be able to implement such requirements in practice adequately, the deployment of nursing or case managers in GP practices would be advisable. Nursing managers could use low‐threshold interventions to assist CHD patients with MHI in the search for a suitable form of psychotherapy and/or drug therapy and to accompany their implementation. They could act as a link between the various care providers and ensure a timely exchange of information and early personal appointments.\\n34\\n\\n\",\n \"MenDis‐CHD provided insights into the needs of CHD patients with and without MHI and their access and barriers to healthcare, clarified important health outcomes for this population and indicated the requirements that a more value‐based care system should focus on to increase quality and efficiency.\\nNonetheless, this study has some limitations. With regard to the qualitative part of the study, we chose to conduct 20 interviews. Looking at the pool of eligible patients, our sample size is limited by the participant diversity and the recruitment location (e.g., recruitment only in one city in Germany and particularly in cardiology practices). Because patients were mainly recruited in outpatient settings, it is possible that they were biased with regard to their preference to talk to a GP about their MHI. Patients may have perceived the outpatient setting as a familiar environment, as there has not necessarily been an inpatient stay recently and could therefore adopt a one‐sided view. Another limitation was that some of the quantitative items in our questionnaire were self‐developed and had no statistical testing of validation. Replication with a larger sample size and greater diversity, either across Germany or internationally, would be advantageous to explore possible generalization effects. Other vulnerable population groups that also have a high prevalence of MHI (e.g., schizophrenia) should be examined to explore specific health outcomes and needs that are important to the particular group of patients. With more data, researchers could build a better basis in clinical practice to facilitate a change towards a value‐based healthcare system.\",\n \"This study found that the needs regarding CHD patients with and without MHI were not adequately satisfied. Furthermore, for CHD patients with MHI, access and barriers to the healthcare system regarding psychotherapeutic and drug therapy were insufficient. CHD patients experienced the cardiac event as life‐changing and had an urgent need to talk about mental problems. Primarily, the GP was named as a partner for conversations about CHD‐related mental problems; however, when the doctor spoke to the patient shortly after the cardiac event, patients felt relieved and coped better with their illness. To improve this, psychological measures such as training of doctors in doctor–patient communication are necessary, as well as improvements in procedural and structural requirements.\",\n \"Samia Peltzer wrote the abstract, introduction, method section, results (incl. analyses) and discussion, designed the figure and tables, references with EndNote, made an overall adaption of the text and included references. Samia Peltzer was a major contributor in writing the manuscript. Ursula Köstler conducted the information for the qualitative analysis, conducted qualitative analysis and summed up the results for the qualitative part. Frank Jessen, Frank Schulz‐Nieswandt and Christian Albus designed and conducted the study as chief scientists. Hendrik Müller, Nadine Scholten, Frank Jessen and Christian Albus critically revised the manuscript. All gave final approval and agreed to be accountable for all aspects of work, ensuring integrity and accuracy.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,"materials-and-methods",null,null,null,null,null,null,null,null,null,null,null,null,"results",null,null,null,null,null,null,null,null,null,null,null,null,"discussion",null,null,null,"conclusions",null],"string":"[\n null,\n null,\n null,\n null,\n null,\n \"materials-and-methods\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n null,\n null,\n null,\n \"conclusions\",\n null\n]"},"keywords":{"kind":"list like","value":["coronary heart disease","healthcare system","mental health issues","mixed‐method approach","value‐base healthcare"],"string":"[\n \"coronary heart disease\",\n \"healthcare system\",\n \"mental health issues\",\n \"mixed‐method approach\",\n \"value‐base healthcare\"\n]"}}},{"rowIdx":389,"cells":{"title":{"kind":"string","value":"Failure pattern and suggestions for target volume delineation of carcinoma showing thymus-like differentiation treated with intensity-modulated radiotherapy."},"pmid":{"kind":"string","value":"36271328"},"background_abstract":{"kind":"string","value":"To review our long-term clinical experience, analyze the failure patterns, and give suggestions for target volume delineation of carcinoma showing thymus-like differentiation (CASTLE) treated with intensity-modulated radiotherapy (IMRT)."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. A total dose of 56-60 Gy in 28-30 fractions was prescribed to patients without residual disease and 66 Gy in 33 fractions for patients with residual or unresectable disease. Survival rates were calculated using the Kaplan-Meier method. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"Among the 30 patients, 12 (40%) received partial resection or biopsy. Lateral lymph node metastasis was observed in 7 (23.3%) patients. During follow-up, regional lymph node recurrence occurred in 2 patients and distant metastasis in 5 patients. With a median follow-up time of 63.5 months, the 5-year local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS) and progression-free survival (PFS) rates were 100, 88.9, 78.9, 93.1 and 78.9%, respectively. For patients with no lateral neck node metastasis, prophylactic radiotherapy for lateral neck nodal regions failed to improve RRFS (p = 0.381) and OS (p = 0.153)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Distant metastasis was the major failure pattern for CASTLE after surgery and IMRT. For patients with no lateral neck node metastasis, the omission of irradiation for lateral neck nodal regions seems to be safe and feasible."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Radiotherapy, Intensity-Modulated","Retrospective Studies","Carcinoma","Radiotherapy Planning, Computer-Assisted","Lymphatic Metastasis"],"string":"[\n \"Humans\",\n \"Radiotherapy, Intensity-Modulated\",\n \"Retrospective Studies\",\n \"Carcinoma\",\n \"Radiotherapy Planning, Computer-Assisted\",\n \"Lymphatic Metastasis\"\n]"},"pmcid":{"kind":"string","value":"9585782"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] Patient characteristics [SUBSECTION] A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\nA total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\n[SUBTITLE] Dosimetric data for IMRT [SUBSECTION] Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\nDose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n[SUBTITLE] Survival outcome [SUBSECTION] The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions\nThe median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions\n[SUBTITLE] Failure patterns and salvage treatment [SUBSECTION] Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\nTwo patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\n[SUBTITLE] Treatment complications [SUBSECTION] The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy\nThe majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy\n[SUBTITLE] Prognostic factors [SUBSECTION] Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\nUnivariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events."},"conclusions_title":{"kind":"string","value":"Conclusion"},"conclusions_text":{"kind":"string","value":"Our long-term results showed that surgery combined with IMRT is an effective treatment for patients with CASTLE. Distant metastasis is the major failure pattern. For patients with no lateral neck node metastasis, the omission of irradiation for lateral neck nodal regions seems to be safe and feasible. Further prospective research is warranted."},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Patients","Intensity-modulate radiotherapy","Patient evaluation","Definition of failure pattern","Statistical methods","Patient characteristics","Dosimetric data for IMRT","Survival outcome","Failure patterns and salvage treatment","Treatment complications","Prognostic factors",""],"string":"[\n \"Background\",\n \"Methods\",\n \"Patients\",\n \"Intensity-modulate radiotherapy\",\n \"Patient evaluation\",\n \"Definition of failure pattern\",\n \"Statistical methods\",\n \"Patient characteristics\",\n \"Dosimetric data for IMRT\",\n \"Survival outcome\",\n \"Failure patterns and salvage treatment\",\n \"Treatment complications\",\n \"Prognostic factors\",\n \"\"\n]"},"other_sections_texts":{"kind":"list like","value":["Carcinoma showing thymus-like differentiation (CASTLE) is a rare malignant tumor of the thyroid or adjacent soft tissue in the neck, which accounts for only 0.1–0.15% of all thyroid cancers [1–4]. Until now, fewer than 100 patients have been reported, and most of them are case reports. Optimal treatment for CASTLE remains uncertain. Surgery is generally recommended as the mainstay of treatment [3, 5]. However, due to the lower location and high possibility of extrathyroid invasion, complete resection can be challenging for patients with locally advanced disease [6]. It is reported that the incidence of lymph node metastasis is 50–69% and 60–80% for extrathyroid invasion [5–9]. Therefore, radiotherapy is often used as part of the treatment and the efficiency of radiotherapy has been proved by several studies [6, 7, 10, 11].\nHowever, there is no consensus on the delineation of target volume for CASTLE since very limited data has been published. Accurate target volume delineation is the premise of intensity-modulated radiotherapy (IMRT). Excessive irradiation will increase treatment-related toxicities like dermatitis and neck fibrosis, while insufficiency of target volume may cause tumor recurrence. Hence, in the present study, we reviewed our 11 years of clinical experience, analyzed the failure patterns, and gave suggestions for target volume delineation of CASTLE treated with IMRT. To the best of our knowledge, this is the first study focusing on target volume delineation for this rare disease.","[SUBTITLE] Patients [SUBSECTION] From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\nFrom April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\n[SUBTITLE] Intensity-modulate radiotherapy [SUBSECTION] IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.\nIMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.\n[SUBTITLE] Patient evaluation [SUBSECTION] Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\nPatients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\n[SUBTITLE] Definition of failure pattern [SUBSECTION] The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\nThe images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\n[SUBTITLE] Statistical methods [SUBSECTION] All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\nAll the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.","From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.","IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.","Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.","The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.","All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.","A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.","Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume","The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions","Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.","The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy","Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"Carcinoma showing thymus-like differentiation (CASTLE) is a rare malignant tumor of the thyroid or adjacent soft tissue in the neck, which accounts for only 0.1–0.15% of all thyroid cancers [1–4]. Until now, fewer than 100 patients have been reported, and most of them are case reports. Optimal treatment for CASTLE remains uncertain. Surgery is generally recommended as the mainstay of treatment [3, 5]. However, due to the lower location and high possibility of extrathyroid invasion, complete resection can be challenging for patients with locally advanced disease [6]. It is reported that the incidence of lymph node metastasis is 50–69% and 60–80% for extrathyroid invasion [5–9]. Therefore, radiotherapy is often used as part of the treatment and the efficiency of radiotherapy has been proved by several studies [6, 7, 10, 11].\\nHowever, there is no consensus on the delineation of target volume for CASTLE since very limited data has been published. Accurate target volume delineation is the premise of intensity-modulated radiotherapy (IMRT). Excessive irradiation will increase treatment-related toxicities like dermatitis and neck fibrosis, while insufficiency of target volume may cause tumor recurrence. Hence, in the present study, we reviewed our 11 years of clinical experience, analyzed the failure patterns, and gave suggestions for target volume delineation of CASTLE treated with IMRT. To the best of our knowledge, this is the first study focusing on target volume delineation for this rare disease.\",\n \"[SUBTITLE] Patients [SUBSECTION] From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\\nFrom April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\\n[SUBTITLE] Intensity-modulate radiotherapy [SUBSECTION] IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\\nIMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\\n[SUBTITLE] Patient evaluation [SUBSECTION] Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\\nPatients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\\n[SUBTITLE] Definition of failure pattern [SUBSECTION] The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\\nThe images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\\n[SUBTITLE] Statistical methods [SUBSECTION] All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\\nAll the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\",\n \"From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\",\n \"IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\",\n \"Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\",\n \"The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\",\n \"All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\",\n \"A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\\n\\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nPatient characteristics\\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\n\\nTreatment procedures and clinical outcome\\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\",\n \"Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\\n\\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\n\\nDose-volume histograms (DVHs) statistics for IMRT\\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\",\n \"The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\\n\\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\\n\\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\\n RT, radiotherapy; LR: lateral neck nodal regions\",\n \"Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\",\n \"The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\\n\\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\\n\\nAcute toxicities during IMRT\\nIMRT, intensity-modulated radiotherapy\",\n \"Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Patients","Intensity-modulate radiotherapy","Patient evaluation","Definition of failure pattern","Statistical methods","Results","Patient characteristics","Dosimetric data for IMRT","Survival outcome","Failure patterns and salvage treatment","Treatment complications","Prognostic factors","Discussion","Conclusion","Electronic supplementary material",""],"string":"[\n \"Background\",\n \"Methods\",\n \"Patients\",\n \"Intensity-modulate radiotherapy\",\n \"Patient evaluation\",\n \"Definition of failure pattern\",\n \"Statistical methods\",\n \"Results\",\n \"Patient characteristics\",\n \"Dosimetric data for IMRT\",\n \"Survival outcome\",\n \"Failure patterns and salvage treatment\",\n \"Treatment complications\",\n \"Prognostic factors\",\n \"Discussion\",\n \"Conclusion\",\n \"Electronic supplementary material\",\n \"\"\n]"},"all_sections_texts":{"kind":"list like","value":["Carcinoma showing thymus-like differentiation (CASTLE) is a rare malignant tumor of the thyroid or adjacent soft tissue in the neck, which accounts for only 0.1–0.15% of all thyroid cancers [1–4]. Until now, fewer than 100 patients have been reported, and most of them are case reports. Optimal treatment for CASTLE remains uncertain. Surgery is generally recommended as the mainstay of treatment [3, 5]. However, due to the lower location and high possibility of extrathyroid invasion, complete resection can be challenging for patients with locally advanced disease [6]. It is reported that the incidence of lymph node metastasis is 50–69% and 60–80% for extrathyroid invasion [5–9]. Therefore, radiotherapy is often used as part of the treatment and the efficiency of radiotherapy has been proved by several studies [6, 7, 10, 11].\nHowever, there is no consensus on the delineation of target volume for CASTLE since very limited data has been published. Accurate target volume delineation is the premise of intensity-modulated radiotherapy (IMRT). Excessive irradiation will increase treatment-related toxicities like dermatitis and neck fibrosis, while insufficiency of target volume may cause tumor recurrence. Hence, in the present study, we reviewed our 11 years of clinical experience, analyzed the failure patterns, and gave suggestions for target volume delineation of CASTLE treated with IMRT. To the best of our knowledge, this is the first study focusing on target volume delineation for this rare disease.","[SUBTITLE] Patients [SUBSECTION] From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\nFrom April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\n[SUBTITLE] Intensity-modulate radiotherapy [SUBSECTION] IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.\nIMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.\n[SUBTITLE] Patient evaluation [SUBSECTION] Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\nPatients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\n[SUBTITLE] Definition of failure pattern [SUBSECTION] The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\nThe images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\n[SUBTITLE] Statistical methods [SUBSECTION] All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\nAll the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.","From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.","IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\n\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\n\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\n PTV-C, planning target volume for CTV.","Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.","The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.","All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.","[SUBTITLE] Patient characteristics [SUBSECTION] A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\nA total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\n[SUBTITLE] Dosimetric data for IMRT [SUBSECTION] Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\nDose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n[SUBTITLE] Survival outcome [SUBSECTION] The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions\nThe median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions\n[SUBTITLE] Failure patterns and salvage treatment [SUBSECTION] Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\nTwo patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\n[SUBTITLE] Treatment complications [SUBSECTION] The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy\nThe majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy\n[SUBTITLE] Prognostic factors [SUBSECTION] Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\nUnivariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.","A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\n\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nPatient characteristics\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\n\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\n\nTreatment procedures and clinical outcome\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.","Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\n\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\n\nDose-volume histograms (DVHs) statistics for IMRT\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume","The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\n\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\n\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\n\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\n RT, radiotherapy; LR: lateral neck nodal regions","Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.","The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\n\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\n\nAcute toxicities during IMRT\nIMRT, intensity-modulated radiotherapy","Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.","CASTLE is an extremely rare malignant tumor, which is thought to arise from ectopic thymus or branchial pouch remnants [14]. The morphological, immunohistochemical, and molecular features of CASTLE are similar to those of thymic carcinoma. The carcinogenesis of CASTLE is the result of combined actions by a series of oncogenes and tumor suppressors. The study from Wang et al. [15] indicated that abnormal expression of p16, Bcl-2, p53, E-cadherin, C-KIT, CA-IX, EGFR, and HER-2 might play a role in the tumorigenesis and development of CASTLE. Patients with HER-2 overexpression showed a worse prognosis, suggesting that HER-2 overexpression may enhance the invasive and metastatic potential of CASTLE. However, the molecular pathological mechanism of CASTLE is still unclear. Further studies are needed to explore the molecular networks of this rare disease.\nAlthough the optimal treatment modality for CASTLE remains uncertain, an increasing number of studies have indicated the important role of radiotherapy in treating CASTLE. Due to the high possibility of extrathyroid invasion and high incidence of lymph node metastasis, adjuvant radiotherapy is often used as part of the treatment [5–9]. And growing evidence has validated the efficiency of adjuvant radiotherapy in recent years [5–7, 10, 11]. Choi et al. [10] reported that adjuvant radiotherapy reduced about 43% of the recurrence for patients with positive lymph nodes. In the study by Gao et al. [6], the median survival time was significantly longer in the surgery and adjuvant radiotherapy group than in the surgery alone group (17.1 vs. 8.8 years, p = 0.034). Moreover, radiotherapy can also play an essential role in tumor control for patients with unresectable disease. Petra et al. [16] recently reported a patient with locally advanced disease which was not suitable for surgery and thus underwent radical IMRT alone. Exciting complete remission was achieved after IMRT at a dose of 70 Gy. In the present study, palliative resection was administered to one of the patients due to the tumor invasion of adjacent great vessels. Then the patient received radical IMRT at a dose of 66 Gy in 33 fractions. Partial remission was achieved. Seven years after IMRT, the patient is still alive with stable disease at our last follow-up. All the above implies the indispensable role of radiotherapy in treating CASTLE.\nThere is no consensus on target volume delineation for CASTLE. Most of the previous studies reported the radiation dose only and with no specification of the target volume. To the best of our knowledge, this is the first study to explore the proper target volume delineation for this rare disease. As we know, neck nodal level VI is the sentinal node for thyroid tumors. Excessive irradiation for the neck can induce dermatitis, especially for the lower neck (as most of the CASTLE arises in the lower part of the thyroid lobe). Severe dermatitis may cause the interruption of radiotherapy and influence the treatment effect. After treatment, it may also cause neck fibrosis, which will seriously affect patients’ long-term quality of life. Hence, in the present study, we compared the effects of prophylactic irradiation and omitting irradiation to lateral neck regions on the survival rates of patients with no lateral neck node metastasis. The results showed that there was no significant difference in RRFS and OS between the two groups, which suggested that it was safe and feasible to omit irradiation to lateral neck regions for patients with no lateral neck node metastasis. Of course, there may be a certain deviation due to the small sample size and few end-point events. Prospective studies are warranted to further verify our conclusion.\nTwo patients had lymph node recurrence, which occurred in the contralateral supraclavicular fossa. Both of the patients underwent R0 resection at first treatment. One of the patients (Pt.1) had positive lymph nodes in level VI and the other patient (Pt.2) had negative lymph node. Pt.1 received adjuvant RT to the tumor bed and neck levels III, IV, and VI. Pt.2 received RT for tumor bed and neck level IV. Recurrences occurred 3 and 5 years respectively after surgery. It’s worth noting that both of the patients suffered synchronous or metachronous lung and mediastinal lymph node metastasis. It seems that the recurrence of the supraclavicular lymph node may be caused by lung and mediastinal lymph node metastasis.\nDistant metastasis is the major failure pattern for CASTLE after surgery and IMRT. However, the role of chemotherapy remains unclear because of the rarity of the disease. Different regimens have been explored, including cisplatin, epirubicin, docetaxel, irinotecan, vincristine, and cyclophosphamide. But the results seem to be heterogeneous. Hanamura et al. [17] reported a good response to platinum-based chemotherapy of a patient with lung metastasis from CASTLE. The author suggested that CASTLE is a chemosensitive tumor, and chemotherapy should be recommended for patients with advanced or metastatic disease. However, in contrast to Hanamura’s study, Roka et al. [18] found no response to three different regimens of a patient with liver metastasis. In the current study, one patient with gross residual disease after surgery received concurrent chemotherapy (docetaxel and cisplatin). However, the chemotherapy was stopped because of the toxicities after one cycle. CR was achieved for the positive lymph node and PR for the primary tumor after IMRT. Unfortunately, the patient suffered lung metastasis 18 months after IMRT. Systematic therapy combined with IMRT may be an effective treatment option for patients with locally advanced diseases. However, the optimal regimen and combination mode needs to be further investigated.\nImmunotherapy is a revolutionary breakthrough in the treatment of cancer. The effectiveness and safeness have been proved in different tumors in the past few years. As to CASTLE, Lorenz et al. [19] reported the first case in 2019. The patient suffered multiple metastatic diseases (lung, mediastinal, hilar, and upper mesenteric lymph nodes and pleura) 10 years after treatment of CASTLE in the parotid gland. The tumor showed a high expression level of PD-L1 (60%), and the PD-L1 inhibitor (pembrolizumab 200 mg every three weeks) was given to the patient. Partial remission was achieved four months after the start of immunotherapy, and treatment-related toxicities were mild and tolerable. The result was exciting. Further research on immunotherapy with or without chemotherapy or radiotherapy for patients with advanced or metastatic CASTLE is warranted.","Our long-term results showed that surgery combined with IMRT is an effective treatment for patients with CASTLE. Distant metastasis is the major failure pattern. For patients with no lateral neck node metastasis, the omission of irradiation for lateral neck nodal regions seems to be safe and feasible. Further prospective research is warranted.","[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1\nBelow is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1","Below is the link to the electronic supplementary material.\n\nSupplementary Material 1\n\nSupplementary Material 1"],"string":"[\n \"Carcinoma showing thymus-like differentiation (CASTLE) is a rare malignant tumor of the thyroid or adjacent soft tissue in the neck, which accounts for only 0.1–0.15% of all thyroid cancers [1–4]. Until now, fewer than 100 patients have been reported, and most of them are case reports. Optimal treatment for CASTLE remains uncertain. Surgery is generally recommended as the mainstay of treatment [3, 5]. However, due to the lower location and high possibility of extrathyroid invasion, complete resection can be challenging for patients with locally advanced disease [6]. It is reported that the incidence of lymph node metastasis is 50–69% and 60–80% for extrathyroid invasion [5–9]. Therefore, radiotherapy is often used as part of the treatment and the efficiency of radiotherapy has been proved by several studies [6, 7, 10, 11].\\nHowever, there is no consensus on the delineation of target volume for CASTLE since very limited data has been published. Accurate target volume delineation is the premise of intensity-modulated radiotherapy (IMRT). Excessive irradiation will increase treatment-related toxicities like dermatitis and neck fibrosis, while insufficiency of target volume may cause tumor recurrence. Hence, in the present study, we reviewed our 11 years of clinical experience, analyzed the failure patterns, and gave suggestions for target volume delineation of CASTLE treated with IMRT. To the best of our knowledge, this is the first study focusing on target volume delineation for this rare disease.\",\n \"[SUBTITLE] Patients [SUBSECTION] From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\\nFrom April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\\n[SUBTITLE] Intensity-modulate radiotherapy [SUBSECTION] IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\\nIMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\\n[SUBTITLE] Patient evaluation [SUBSECTION] Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\\nPatients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\\n[SUBTITLE] Definition of failure pattern [SUBSECTION] The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\\nThe images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\\n[SUBTITLE] Statistical methods [SUBSECTION] All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\\nAll the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\",\n \"From April 2008 to May 2019, 30 patients with CASTLE treated by postoperative or radical IMRT in our center were retrospectively reviewed. We collected clinicopathological data, treatment procedures, and clinical outcomes. The patterns of treatment failure were also analyzed. The study was approved by the Institutional Review Board of Fudan University Shanghai Cancer Center. Informed consent was obtained from all patients.\",\n \"IMRT was started 4 to 6 weeks after surgery or early after diagnosis. The techniques of IMRT were detailed in our previously published data [7]. Briefly, a thermoplastic mask of the head and shoulder was used for patient immobilization. Patients received computed tomography (CT) simulation at 5 mm thickness of the head and neck region in the supine position. Image fusion of magnetic resonance imaging (MRI) and CT was recommended for target volume delineation. The gross tumor volume (GTV) was defined as all primary gross tumors and involved lymph nodes determined by imaging and clinical findings. The clinical target volume (CTV) included the GTV or tumor bed plus a 5 to 10 mm margin to encompass any microscopic extension. The planning target volume (PTV) was defined as the CTV plus a 3 to 5 mm margin to encompass setup error. Neck nodal level VI was conventionally included in the CTV. If there was no positive lymph node, neck nodal levels II-V were not included in the CTV. Examples of target delineation were showed in Fig. 1. Positive lymph nodes were defined as follows: (1) pathologically diagnosed; (2) minimum axial diameter ≥ 1 cm; (3) extranodal extension or circular enhancement. The prescribed dose was 60 Gy for patients without residual disease and 66 Gy for patients with residual or unresectable disease. Conventional fractionation (2 Gy per fraction, one fraction per day, five days per week) was used.\\n\\nFig. 1Examples of target volume delineation for carcinoma showing thymus-like differentiation GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume; PTV-C, planning target volume for CTV.\\n\\nExamples of target volume delineation for carcinoma showing thymus-like differentiation\\n GTV, gross tumor volume; PTV-G, planning target volume for GTV; CTV, clinical target volume;\\n PTV-C, planning target volume for CTV.\",\n \"Patients were assessed weekly during IMRT. After IMRT, patients were follow-up every three months in the first two years, every six months during the years 3–5, and annually after that. Follow-up assessments after treatment included examination of the neck, thyroid function tests, and ultrasound or MRI for the thyroid and neck. Chest CT scan, and ultrasound or CT of the abdomen were performed every 6–12 months. Additional tests were performed when clinically indicated.\",\n \"The images of MRI or CT scans obtained at the time of recurrence were transferred to the pretreatment planning CT. The dose-volume histogram (DVH) was used to calculate the radiation dose received by the recurrent tumor (GTVrecur) region. The recurrences were classified into five types based on combined spatial and dosimetric criteria [12, 13]:\\nType A (central high dose): the mapped centroid of GTVrecur originates in high dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType B (peripheral high dose): the mapped centroid of GTVrecur originates in high dose PTV, and < 95% of GTVrecur was within the 95% isodose (high dose PTV);\\nType C (central elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and ≥ 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType D (peripheral elective dose): the mapped centroid of GTVrecur originates in lower dose PTV, and < 95% of GTVrecur was within the 95% isodose (lower dose PTV);\\nType E (extraneous dose): the mapped centroid of GTVrecur originates outside all PTVs.\",\n \"All the statistical analysis was performed using the Statistical Package for Social Sciences (SPSS version 24.0) software. The survival rates were calculated from the day of the first treatment. The Kaplan-Meier method was used to calculate the local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) rates. Treatment-related toxicities were graded by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0.\",\n \"[SUBTITLE] Patient characteristics [SUBSECTION] A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\\n\\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nPatient characteristics\\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\n\\nTreatment procedures and clinical outcome\\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\\nA total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\\n\\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nPatient characteristics\\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\n\\nTreatment procedures and clinical outcome\\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\\n[SUBTITLE] Dosimetric data for IMRT [SUBSECTION] Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\\n\\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\n\\nDose-volume histograms (DVHs) statistics for IMRT\\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\nDose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\\n\\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\n\\nDose-volume histograms (DVHs) statistics for IMRT\\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\n[SUBTITLE] Survival outcome [SUBSECTION] The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\\n\\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\\n\\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\\n RT, radiotherapy; LR: lateral neck nodal regions\\nThe median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\\n\\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\\n\\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\\n RT, radiotherapy; LR: lateral neck nodal regions\\n[SUBTITLE] Failure patterns and salvage treatment [SUBSECTION] Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\\nTwo patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\\n[SUBTITLE] Treatment complications [SUBSECTION] The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\\n\\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\\n\\nAcute toxicities during IMRT\\nIMRT, intensity-modulated radiotherapy\\nThe majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\\n\\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\\n\\nAcute toxicities during IMRT\\nIMRT, intensity-modulated radiotherapy\\n[SUBTITLE] Prognostic factors [SUBSECTION] Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\\nUnivariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\",\n \"A total of 30 patients were enrolled in this study. Patients’ characteristics were summarized in Table 1. The median age was 53 years (range 37–61 years). The ratio of male to female was 1:1. Lymph node metastasis was found in 50% of the patients, among which 53.4% were central and 46.6% were lateral. Tumor extension to adjacent organs, including recurrent laryngeal nerve, great vessels, muscles, esophagus, or trachea was found in 21 (70%) patients. About 53% of the patients received complete resection, and 40% received partial resection or biopsy only. Two patients received cisplatin-based concurrent chemotherapy. After treatment, complete remission (CR) was achieved in 26 (86.7%) patients, and partial remission (PR) in 4 (13.3%) patients. Treatment procedures are detailed in Table 2.\\n\\nTable 1Patient characteristicsNo. (%) of patientsTotal30Median age (range)53 (37–61)GenderMale15 (50)Female15 (50)Tumor locationLeft lobe18 (60)Right lobe11 (36.7)Isthmus1 (3.3)Tumor size (cm)≤ 23 (10)2–415 (50)> 47 (23.3)Unknown5 (16.7)Lymph node metastasisPresent15 (50.0)Absent14 (46.6)Unknown1 (3.3)Location of Lymph node(n = 15)Lateral7 (46.6)Central8 (53.4)Tumor extensionPresent21 (70)Absent7 (23.3)Unknown2 (6.7)STE (n = 21)RLN8 (38.1)Muscles6 (28.6)Esophagus3 (14.3)Great vessels4 (19)Trachea2 (9.5)Thymus1 (4.7)Parathyroid gland1 (4.7)Abbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nPatient characteristics\\nAbbreviations: STE, site of tumor extension; RLN, recurrent laryngeal nerve\\n\\nTable 2Treatment procedures and clinical outcomeNo. (%) of patientsType of surgeryR016 (53.3)R12 (6.6)R29 (30)Biopsy3 (10)IMRT dose (Gy)Median (range)60 (56–66)IMRT duration (days)Median (range)44.5 (39–67)Systemic therapyYes vs. No3 (10) vs. 27 (90)Follow-up time (months)Median (range)63.5 (21–161)Treatment responseCR26 (86.7)PR4 (13.3)Local recurrencePresent2 (93.3)Absent28 (6.7)Distant metastasisPresent5 (83.3)Absent25 (16.7)Abbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\n\\nTreatment procedures and clinical outcome\\nAbbreviations: R0, no residual tumor; R1, microscopic residual tumor; R2, macroscopic residual tumor; IMRT, intensity-modulated radiotherapy; CR, complete remission; PR, partial remission\\nPathology and immunohistochemical (IHC) data were available for 29 out of 30 patients. Among these, 93.1% (27/29) expressed CD5, and 72.4% (21/29) expressed CD117 totally or partially. Instead, 72.4% (21/29) of the cases were negatively expressed for thyroid transcription factor 1 (TTF-1), which was the marker of thyroid follicular cells.\",\n \"Dose-volume histogram (DVH) statistics were showed in Table 3. The average volume of GTV and CTV were 131.3 cc (21.2-387.2) and 323 cc （92.4-1408.2）, respectively. Rates of dose coverage were excellent for the target volume. The volume receiving less than 95% of the prescribed dose was 0.3% for GTV and 1.2% for CTV. The mean dose was 67.8 Gy for GTV and 62.7 Gy for CTV.\\n\\nTable 3Dose-volume histograms (DVHs) statistics for IMRTGTVAverage (range)CTVAverage (range)Volume (cc)131.3 (21.2-387.2)323 (92.4-1408.2)Maximum dose (Gy)70.9 (69.2–72.5)67.6 (63.5–72.5)Mean dose (Gy)67.8 (65.9–68.8)62.7 (57.7–66.8)Minimum dose (Gy)56.8 (9.9–65.6)42.1 (38.2–56.3)% volume receiving < 95% of the prescribed dose0.3 (0-1.7)1.2 (0–12)% volume receiving ≥ 100% of the prescribed dose92.1(47.5–100)94.0(77.6–99.7)% volume receiving ≥ 110% of the prescribed dose015.8 (0-67.1)IMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\\n\\nDose-volume histograms (DVHs) statistics for IMRT\\nIMRT, intensity-modulated radiotherapy; GTV, gross tumor volume; CTV, clinical target volume\",\n \"The median follow-up time was 63.5 months (range 21–161 months). The 5-year LRFS, RRFS, DMFS, OS, and PFS were 100, 88.9, 78.9, 93.1, and 78.9%, respectively (Fig. 2). In the subgroup analysis, we divided patients with no lateral neck node metastasis (23 patients) into two groups: group A (6 patients) with prophylactic irradiation for lateral neck nodal regions (levels II-V), while group B (17 patients) without prophylactic irradiation for lateral neck nodal regions. The results showed that there was no significant difference in RRFS (p = 0.381) and OS (p = 0.153) between the two groups (Fig. 3). Patient characteristics for groups A and B were shown in Supplementary Table 1.\\n\\nFig. 2Kaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nKaplan-Meier curves showing local recurrence-free survival (LRFS), regional recurrence-free survival (RRFS), distant metastases-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) for patients with carcinoma showing thymus-like differentiation\\n\\nFig. 3Kaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions RT, radiotherapy; LR: lateral neck nodal regions\\n\\nKaplan-Meier estimate of (a) regional recurrence-free survival (RRFS), and (b) overall survival (OS) between patients with or without irradiation for lateral neck nodal regions\\n RT, radiotherapy; LR: lateral neck nodal regions\",\n \"Two patients had regional lymph node recurrence. All the recurrences were in the contralateral supraclavicular fossa and were defined as type E (extraneous) failures. Distant metastasis was found in 5 patients. Common sites for distant metastasis were lung (3 patients), bone (1 patient), and liver (1 patient). Two patients suffered both regional and distant failure.\\nSalvage treatment, including surgery, radiotherapy, and systemic therapy, was given to patients with recurrence or distant metastases. If acceptable, clinical trials were the first choice. For patients with oligometastatic or oligoprogressive tumors, SBRT was implemented.\",\n \"The majority of acute toxicities related to IMRT were grade 1–2. Grade 3 mucositis or dermatitis occurred only in 2 patients. No grade 4 toxicities were observed. Details were showed in Table 4. Three patients suffered treatment interruption due to other diseases, including aspiration pneumonia and herpes zoster. After appropriate treatment, they continued and finished IMRT. There was no late toxicity (≥ grade 2) related to IMRT, such as radiation pneumonitis, neck fibrosis, or esophageal stenosis.\\n\\nTable 4Acute toxicities during IMRTToxicitiesNo. of patients by toxicity grade (%)Grade 1Grade 2Grade 3Grade 4Grade 5Mucositis9 (30)20 (66.7)1 (3.3)00Dermatitis25 (83.3)4 (13.3)1 (3.3)00Skin pigmentation26 (86.7)0000Anemia4 (13.3)1 (3.3)000Leukopenia5 (16.7)1 (3.3)000Thrombocytopenia3 (10)0000IMRT, intensity-modulated radiotherapy\\n\\nAcute toxicities during IMRT\\nIMRT, intensity-modulated radiotherapy\",\n \"Univariate analysis, taking age, gender, tumor size, tumor extension, surgery type, and lymph node metastasis as prognostic factors for RRFS, DMFS, OS, and PFS, was performed. Univariate analysis showed that surgery type (p = 0.026, 0.000, 0.026) and lateral neck node metastasis (p = 0.024, 0.009, 0.024) were potential prognostic factors for DMFS, OS and PFS (Supplementary Table 2). However, the multivariate analysis failed to further validate the differences (Supplementary Table 3). We attributed it to the relatively small sample size and few end-point events.\",\n \"CASTLE is an extremely rare malignant tumor, which is thought to arise from ectopic thymus or branchial pouch remnants [14]. The morphological, immunohistochemical, and molecular features of CASTLE are similar to those of thymic carcinoma. The carcinogenesis of CASTLE is the result of combined actions by a series of oncogenes and tumor suppressors. The study from Wang et al. [15] indicated that abnormal expression of p16, Bcl-2, p53, E-cadherin, C-KIT, CA-IX, EGFR, and HER-2 might play a role in the tumorigenesis and development of CASTLE. Patients with HER-2 overexpression showed a worse prognosis, suggesting that HER-2 overexpression may enhance the invasive and metastatic potential of CASTLE. However, the molecular pathological mechanism of CASTLE is still unclear. Further studies are needed to explore the molecular networks of this rare disease.\\nAlthough the optimal treatment modality for CASTLE remains uncertain, an increasing number of studies have indicated the important role of radiotherapy in treating CASTLE. Due to the high possibility of extrathyroid invasion and high incidence of lymph node metastasis, adjuvant radiotherapy is often used as part of the treatment [5–9]. And growing evidence has validated the efficiency of adjuvant radiotherapy in recent years [5–7, 10, 11]. Choi et al. [10] reported that adjuvant radiotherapy reduced about 43% of the recurrence for patients with positive lymph nodes. In the study by Gao et al. [6], the median survival time was significantly longer in the surgery and adjuvant radiotherapy group than in the surgery alone group (17.1 vs. 8.8 years, p = 0.034). Moreover, radiotherapy can also play an essential role in tumor control for patients with unresectable disease. Petra et al. [16] recently reported a patient with locally advanced disease which was not suitable for surgery and thus underwent radical IMRT alone. Exciting complete remission was achieved after IMRT at a dose of 70 Gy. In the present study, palliative resection was administered to one of the patients due to the tumor invasion of adjacent great vessels. Then the patient received radical IMRT at a dose of 66 Gy in 33 fractions. Partial remission was achieved. Seven years after IMRT, the patient is still alive with stable disease at our last follow-up. All the above implies the indispensable role of radiotherapy in treating CASTLE.\\nThere is no consensus on target volume delineation for CASTLE. Most of the previous studies reported the radiation dose only and with no specification of the target volume. To the best of our knowledge, this is the first study to explore the proper target volume delineation for this rare disease. As we know, neck nodal level VI is the sentinal node for thyroid tumors. Excessive irradiation for the neck can induce dermatitis, especially for the lower neck (as most of the CASTLE arises in the lower part of the thyroid lobe). Severe dermatitis may cause the interruption of radiotherapy and influence the treatment effect. After treatment, it may also cause neck fibrosis, which will seriously affect patients’ long-term quality of life. Hence, in the present study, we compared the effects of prophylactic irradiation and omitting irradiation to lateral neck regions on the survival rates of patients with no lateral neck node metastasis. The results showed that there was no significant difference in RRFS and OS between the two groups, which suggested that it was safe and feasible to omit irradiation to lateral neck regions for patients with no lateral neck node metastasis. Of course, there may be a certain deviation due to the small sample size and few end-point events. Prospective studies are warranted to further verify our conclusion.\\nTwo patients had lymph node recurrence, which occurred in the contralateral supraclavicular fossa. Both of the patients underwent R0 resection at first treatment. One of the patients (Pt.1) had positive lymph nodes in level VI and the other patient (Pt.2) had negative lymph node. Pt.1 received adjuvant RT to the tumor bed and neck levels III, IV, and VI. Pt.2 received RT for tumor bed and neck level IV. Recurrences occurred 3 and 5 years respectively after surgery. It’s worth noting that both of the patients suffered synchronous or metachronous lung and mediastinal lymph node metastasis. It seems that the recurrence of the supraclavicular lymph node may be caused by lung and mediastinal lymph node metastasis.\\nDistant metastasis is the major failure pattern for CASTLE after surgery and IMRT. However, the role of chemotherapy remains unclear because of the rarity of the disease. Different regimens have been explored, including cisplatin, epirubicin, docetaxel, irinotecan, vincristine, and cyclophosphamide. But the results seem to be heterogeneous. Hanamura et al. [17] reported a good response to platinum-based chemotherapy of a patient with lung metastasis from CASTLE. The author suggested that CASTLE is a chemosensitive tumor, and chemotherapy should be recommended for patients with advanced or metastatic disease. However, in contrast to Hanamura’s study, Roka et al. [18] found no response to three different regimens of a patient with liver metastasis. In the current study, one patient with gross residual disease after surgery received concurrent chemotherapy (docetaxel and cisplatin). However, the chemotherapy was stopped because of the toxicities after one cycle. CR was achieved for the positive lymph node and PR for the primary tumor after IMRT. Unfortunately, the patient suffered lung metastasis 18 months after IMRT. Systematic therapy combined with IMRT may be an effective treatment option for patients with locally advanced diseases. However, the optimal regimen and combination mode needs to be further investigated.\\nImmunotherapy is a revolutionary breakthrough in the treatment of cancer. The effectiveness and safeness have been proved in different tumors in the past few years. As to CASTLE, Lorenz et al. [19] reported the first case in 2019. The patient suffered multiple metastatic diseases (lung, mediastinal, hilar, and upper mesenteric lymph nodes and pleura) 10 years after treatment of CASTLE in the parotid gland. The tumor showed a high expression level of PD-L1 (60%), and the PD-L1 inhibitor (pembrolizumab 200 mg every three weeks) was given to the patient. Partial remission was achieved four months after the start of immunotherapy, and treatment-related toxicities were mild and tolerable. The result was exciting. Further research on immunotherapy with or without chemotherapy or radiotherapy for patients with advanced or metastatic CASTLE is warranted.\",\n \"Our long-term results showed that surgery combined with IMRT is an effective treatment for patients with CASTLE. Distant metastasis is the major failure pattern. For patients with no lateral neck node metastasis, the omission of irradiation for lateral neck nodal regions seems to be safe and feasible. Further prospective research is warranted.\",\n \"[SUBTITLE] [SUBSECTION] Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\\nBelow is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\",\n \"Below is the link to the electronic supplementary material.\\n\\nSupplementary Material 1\\n\\nSupplementary Material 1\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,"results",null,null,null,null,null,null,"discussion","conclusion","supplementary-material",null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\",\n \"conclusion\",\n \"supplementary-material\",\n null\n]"},"keywords":{"kind":"list like","value":["Intensity-modulated radiotherapy","IMRT","Radiotherapy","Carcinoma showing thymus-like differentiation","CASTLE"],"string":"[\n \"Intensity-modulated radiotherapy\",\n \"IMRT\",\n \"Radiotherapy\",\n \"Carcinoma showing thymus-like differentiation\",\n \"CASTLE\"\n]"}}},{"rowIdx":390,"cells":{"title":{"kind":"string","value":"Sulbactam combined with tigecycline improves outcomes in patients with severe multidrug-resistant Acinetobacter baumannii pneumonia."},"pmid":{"kind":"string","value":"36271362"},"background_abstract":{"kind":"string","value":"The purpose of this study was to review the treatment plan of patients with multidrug-resistant Acinetobacter baumannii (MDR-AB) pneumonia and analyze the factors associated with patient deaths and the medication regimen."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We collected 1,823 qualified respiratory specimens that were culture-positive for MDR-AB. 166 patients confirmed to have hospital-acquired MDR-AB pneumonia were selected as the research subjects. The differing clinical characteristics and treatment interventions between the surviving group and death group within 28 days were analyzed."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"The mortality rate was high for those aged > 75 years (p = 0.001). Patients who underwent invasive catheter placement (p < 0.001) and mechanical ventilation (p = 0.046) had a higher mortality rate. Combination therapy with tigecycline can reduce the mortality rate (p < 0.001) of MDR-AB pneumonia in patients with carbapenem-resistant AB(CRAB). Combination therapy with sulbactam was shown to reduce the mortality rate (p < 0.001), and high-dose sulbactam (> 3 g/day) might be better than low-dose sulbactam (≤ 3 g/day)."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Reducing the time of invasive catheter placement and mechanical ventilation in patients in the intensive care unit (ICU), antimicrobial treatment, combined with tigecycline and sulbactam, might help reduce the mortality rate in patients with severe MDR-AB hospital-acquired pneumonia."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Humans","Acinetobacter baumannii","Sulbactam","Tigecycline","Acinetobacter Infections","Drug Resistance, Multiple, Bacterial","Carbapenems","Anti-Bacterial Agents","Anti-Infective Agents","Pneumonia","Retrospective Studies","Microbial Sensitivity Tests"],"string":"[\n \"Humans\",\n \"Acinetobacter baumannii\",\n \"Sulbactam\",\n \"Tigecycline\",\n \"Acinetobacter Infections\",\n \"Drug Resistance, Multiple, Bacterial\",\n \"Carbapenems\",\n \"Anti-Bacterial Agents\",\n \"Anti-Infective Agents\",\n \"Pneumonia\",\n \"Retrospective Studies\",\n \"Microbial Sensitivity Tests\"\n]"},"pmcid":{"kind":"string","value":"9585752"},"background_title":{"kind":"null"},"background_text":{"kind":"null"},"methods_title":{"kind":"null"},"methods_text":{"kind":"null"},"results_title":{"kind":"string","value":"Results"},"results_text":{"kind":"string","value":"[SUBTITLE] General clinical conditions [SUBSECTION] Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\nOf the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n[SUBTITLE] Characteristics of Acinetobacter baumannii strains(abs) [SUBSECTION] Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem\nOf the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem\n[SUBTITLE] Antibiotic medication [SUBSECTION] All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\nAll patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n[SUBTITLE] Tigecycline [SUBSECTION] Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\nAmong the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\n[SUBTITLE] Carbapenem [SUBSECTION] There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\nThere were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\n[SUBTITLE] Sulbactam [SUBSECTION] From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\nFrom the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n[SUBTITLE] Colistin [SUBSECTION] There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\nThere were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin."},"conclusions_title":{"kind":"null"},"conclusions_text":{"kind":"null"},"other_sections_titles":{"kind":"list like","value":["Background","Methods","Subjects","Method for detecting carbapenemase-producing genotype","Clinical information","Treatment","Statistical methods","General clinical conditions","Characteristics of Acinetobacter baumannii strains(abs)","Antibiotic medication","Tigecycline","Carbapenem","Sulbactam","Colistin"],"string":"[\n \"Background\",\n \"Methods\",\n \"Subjects\",\n \"Method for detecting carbapenemase-producing genotype\",\n \"Clinical information\",\n \"Treatment\",\n \"Statistical methods\",\n \"General clinical conditions\",\n \"Characteristics of Acinetobacter baumannii strains(abs)\",\n \"Antibiotic medication\",\n \"Tigecycline\",\n \"Carbapenem\",\n \"Sulbactam\",\n \"Colistin\"\n]"},"other_sections_texts":{"kind":"list like","value":["Acinetobacter baumannii (AB) is a common pathogen that causes opportunistic infections in hospitals. According to the China Antimicrobial Surveillance Network (CHINET) continuous drug resistance monitoring program, AB is the second most common gram-negative bacterium isolated from respiratory specimens and accounted for about 17.07% of infections in 2020. Important risk factors for AB infection include admission to the ICU, impaired immune function, receiving broad-spectrum antibacterial drugs, and mechanical ventilation [1,2]. Since AB has a high rate of resistance to antimicrobial drugs, the control of multi-drug resistant AB (MDR-AB) is currently one of the main goals of nosocomial infection prevention and control.\nIn Asia, the sensitivity of AB to carbapenems were less than 27% [3]. As such, a number of guidelines recommend the combination of polymyxin, tigecycline, carbapenems, sulbactam, or other antibacterial drugs for the treatment of MDR-AB pneumonia in China or abroad [4, 5]. However, the sensitivity of AB to certain drugs differs in different regions. Thus, it is necessary to design an appropriate combination drug regimen according to the local drug resistance situation and common guidelines and recommendations.\nIt is a challenge for clinicians to choose antibacterial drugs when trying to control severe MDR-AB infection, especially carbapenemase-resistant Acinetobacter baumannii (CRAB). According to an in vitro drug sensitivity test, the resistance rate of AB to carbapenems ranged from 62–100% [3], and the resistance rate to cefoperazone/sulbactam was 46.3% [6]. Antibiotics containing sulbactam were the main drugs for the treatment of Acinetobacter infection in China, partly because of the economic burden. The conventional dose is 4 g/d for MDR-AB, but it can be increased to 6.0 g/d or even 8.0 g/d according to expert consensus on the diagnosis, treatment, prevention, and control of AB infection in China [4]. In addition to sulbactam, tigecycline is the most used for the control of hospital-acquired MDR-AB, but its clinical efficacy is controversial [7]. Studies have shown that even if AB was sensitive to tigecycline in vitro, it could not improve the prognosis of patients with MDR-AB. The main objective of this study was to investigate feasible and effective treatment options for patients with severe MDR-AB pneumonia, from a clinical perspective. Here, we retrospectively identified factors influencing mortality in patients with severe MDR-AB pneumonia and appropriate antibiotic regimens, especially the combined effect of sulbactam and tigecycline.","[SUBTITLE] Subjects [SUBSECTION] We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\nWe enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n[SUBTITLE] Method for detecting carbapenemase-producing genotype [SUBSECTION] The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\nThe carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\n[SUBTITLE] Clinical information [SUBSECTION] For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\nFor each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\n[SUBTITLE] Treatment [SUBSECTION] Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\nEach treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n[SUBTITLE] Statistical methods [SUBSECTION] SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\nSPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.","We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii","The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.","For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.","Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.","SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.","Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii","Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem","All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines","Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.","There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).","From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii","There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin."],"string":"[\n \"Acinetobacter baumannii (AB) is a common pathogen that causes opportunistic infections in hospitals. According to the China Antimicrobial Surveillance Network (CHINET) continuous drug resistance monitoring program, AB is the second most common gram-negative bacterium isolated from respiratory specimens and accounted for about 17.07% of infections in 2020. Important risk factors for AB infection include admission to the ICU, impaired immune function, receiving broad-spectrum antibacterial drugs, and mechanical ventilation [1,2]. Since AB has a high rate of resistance to antimicrobial drugs, the control of multi-drug resistant AB (MDR-AB) is currently one of the main goals of nosocomial infection prevention and control.\\nIn Asia, the sensitivity of AB to carbapenems were less than 27% [3]. As such, a number of guidelines recommend the combination of polymyxin, tigecycline, carbapenems, sulbactam, or other antibacterial drugs for the treatment of MDR-AB pneumonia in China or abroad [4, 5]. However, the sensitivity of AB to certain drugs differs in different regions. Thus, it is necessary to design an appropriate combination drug regimen according to the local drug resistance situation and common guidelines and recommendations.\\nIt is a challenge for clinicians to choose antibacterial drugs when trying to control severe MDR-AB infection, especially carbapenemase-resistant Acinetobacter baumannii (CRAB). According to an in vitro drug sensitivity test, the resistance rate of AB to carbapenems ranged from 62–100% [3], and the resistance rate to cefoperazone/sulbactam was 46.3% [6]. Antibiotics containing sulbactam were the main drugs for the treatment of Acinetobacter infection in China, partly because of the economic burden. The conventional dose is 4 g/d for MDR-AB, but it can be increased to 6.0 g/d or even 8.0 g/d according to expert consensus on the diagnosis, treatment, prevention, and control of AB infection in China [4]. In addition to sulbactam, tigecycline is the most used for the control of hospital-acquired MDR-AB, but its clinical efficacy is controversial [7]. Studies have shown that even if AB was sensitive to tigecycline in vitro, it could not improve the prognosis of patients with MDR-AB. The main objective of this study was to investigate feasible and effective treatment options for patients with severe MDR-AB pneumonia, from a clinical perspective. Here, we retrospectively identified factors influencing mortality in patients with severe MDR-AB pneumonia and appropriate antibiotic regimens, especially the combined effect of sulbactam and tigecycline.\",\n \"[SUBTITLE] Subjects [SUBSECTION] We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\nWe enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n[SUBTITLE] Method for detecting carbapenemase-producing genotype [SUBSECTION] The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\\nThe carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\\n[SUBTITLE] Clinical information [SUBSECTION] For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\\nFor each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\\n[SUBTITLE] Treatment [SUBSECTION] Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\nEach treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n[SUBTITLE] Statistical methods [SUBSECTION] SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\\nSPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\",\n \"We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\",\n \"The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\",\n \"For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\",\n \"Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\",\n \"SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\",\n \"Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\\n\\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nClinical data of patients with MDR-AB pneumonia\\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\",\n \"Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\\n\\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\\n\\nThe susceptibility results of Abs in vitro\\n*Carbapenems including Imipenem/Meropenem/Doripenem\",\n \"All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\\n\\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\n\\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\",\n \"Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\",\n \"There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\",\n \"From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\\n\\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\",\n \"There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\"\n]"},"other_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,null,null,null,null,null,null,null],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n null\n]"},"all_sections_titles":{"kind":"list like","value":["Background","Methods","Subjects","Method for detecting carbapenemase-producing genotype","Clinical information","Treatment","Statistical methods","Results","General clinical conditions","Characteristics of Acinetobacter baumannii strains(abs)","Antibiotic medication","Tigecycline","Carbapenem","Sulbactam","Colistin","Discussion"],"string":"[\n \"Background\",\n \"Methods\",\n \"Subjects\",\n \"Method for detecting carbapenemase-producing genotype\",\n \"Clinical information\",\n \"Treatment\",\n \"Statistical methods\",\n \"Results\",\n \"General clinical conditions\",\n \"Characteristics of Acinetobacter baumannii strains(abs)\",\n \"Antibiotic medication\",\n \"Tigecycline\",\n \"Carbapenem\",\n \"Sulbactam\",\n \"Colistin\",\n \"Discussion\"\n]"},"all_sections_texts":{"kind":"list like","value":["Acinetobacter baumannii (AB) is a common pathogen that causes opportunistic infections in hospitals. According to the China Antimicrobial Surveillance Network (CHINET) continuous drug resistance monitoring program, AB is the second most common gram-negative bacterium isolated from respiratory specimens and accounted for about 17.07% of infections in 2020. Important risk factors for AB infection include admission to the ICU, impaired immune function, receiving broad-spectrum antibacterial drugs, and mechanical ventilation [1,2]. Since AB has a high rate of resistance to antimicrobial drugs, the control of multi-drug resistant AB (MDR-AB) is currently one of the main goals of nosocomial infection prevention and control.\nIn Asia, the sensitivity of AB to carbapenems were less than 27% [3]. As such, a number of guidelines recommend the combination of polymyxin, tigecycline, carbapenems, sulbactam, or other antibacterial drugs for the treatment of MDR-AB pneumonia in China or abroad [4, 5]. However, the sensitivity of AB to certain drugs differs in different regions. Thus, it is necessary to design an appropriate combination drug regimen according to the local drug resistance situation and common guidelines and recommendations.\nIt is a challenge for clinicians to choose antibacterial drugs when trying to control severe MDR-AB infection, especially carbapenemase-resistant Acinetobacter baumannii (CRAB). According to an in vitro drug sensitivity test, the resistance rate of AB to carbapenems ranged from 62–100% [3], and the resistance rate to cefoperazone/sulbactam was 46.3% [6]. Antibiotics containing sulbactam were the main drugs for the treatment of Acinetobacter infection in China, partly because of the economic burden. The conventional dose is 4 g/d for MDR-AB, but it can be increased to 6.0 g/d or even 8.0 g/d according to expert consensus on the diagnosis, treatment, prevention, and control of AB infection in China [4]. In addition to sulbactam, tigecycline is the most used for the control of hospital-acquired MDR-AB, but its clinical efficacy is controversial [7]. Studies have shown that even if AB was sensitive to tigecycline in vitro, it could not improve the prognosis of patients with MDR-AB. The main objective of this study was to investigate feasible and effective treatment options for patients with severe MDR-AB pneumonia, from a clinical perspective. Here, we retrospectively identified factors influencing mortality in patients with severe MDR-AB pneumonia and appropriate antibiotic regimens, especially the combined effect of sulbactam and tigecycline.","[SUBTITLE] Subjects [SUBSECTION] We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\nWe enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n[SUBTITLE] Method for detecting carbapenemase-producing genotype [SUBSECTION] The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\nThe carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\n[SUBTITLE] Clinical information [SUBSECTION] For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\nFor each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\n[SUBTITLE] Treatment [SUBSECTION] Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\nEach treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n[SUBTITLE] Statistical methods [SUBSECTION] SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\nSPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.","We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\n\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii","The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.","For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.","Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\n\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.","SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.","[SUBTITLE] General clinical conditions [SUBSECTION] Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\nOf the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n[SUBTITLE] Characteristics of Acinetobacter baumannii strains(abs) [SUBSECTION] Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem\nOf the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem\n[SUBTITLE] Antibiotic medication [SUBSECTION] All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\nAll patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n[SUBTITLE] Tigecycline [SUBSECTION] Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\nAmong the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\n[SUBTITLE] Carbapenem [SUBSECTION] There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\nThere were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\n[SUBTITLE] Sulbactam [SUBSECTION] From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\nFrom the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n[SUBTITLE] Colistin [SUBSECTION] There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\nThere were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.","Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\n\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\n\nClinical data of patients with MDR-AB pneumonia\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii","Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\n\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\n\nThe susceptibility results of Abs in vitro\n*Carbapenems including Imipenem/Meropenem/Doripenem","All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\n\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\n\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines","Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.","There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).","From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\n\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\n\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii","There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.","MDR-AB increases the mortality rate of nosocomial pneumonia. Our study found that admission to the intensive care unit, age > 75 years, invasive catheterization, and mechanical ventilation increased the mortality of patients with severe MDR-AB pneumonia, which was consistent with the results of other studies [9]. Early removal of invasive catheters and reducing the time on mechanical ventilation might reduce the mortality of critically ill patients diagnosed with severe MDR-AB pneumonia. Table 1 shows that there were age differences in clinical outcomes in the study population, but no differences by gender. This may be related to factors such as more complications, more co-morbidities, difficulty in weaning from ventilator, and longer ICU stay in older adults with severe pneumonia.\nIn our study, the main drugs that affect the mortality of patients with severe MDR-AB pneumonia or CRAB pneumonia included tigecycline, carbapenem, and sulbactam. Although Abs had good sensitivity to colistin in vitro, the MIC of colistin was less than 1 in China. But colistin was not widely used in the treatment of CRAB because of its high cost and not being covered by general health insurance in our country. There were only 2 patients were treated with colistin for CRAB, and they could not be included in the study due to uncertainty of clinical data and co-infection of fungal.\nAlthough tigecycline was recommended as a first-line drug for CRAB pneumonia, there were different opinions on whether its use was beneficial to reduce mortality. The resistance rate of tigecycline in vitro was low in China (about 2.9%) [6], which ranges from 0.2 to 74.2% in other countries and regions [3]. The concentration of tigecycline was higher in tissue, and it has suitable antibacterial activity and was safe for use against local tissue infections. Therefore, it has been theorized that tigecycline was an effective antibiotic for the treatment of CRAB and can reduce mortality [10. A previous meta-analysis showed that patients treated with tigecycline for CRAB infection had a higher mortality rate, longer hospital stay, and lower microbial clearance rate as compared with other drugs [11]. Another prospective study demonstrated that tigecycline increases the mortality rate in patients with CRAB infection when the tigecycline MIC > 2 mg/L [12]. The results of our study showed that tigecycline could significantly reduce the mortality rate of hospital-acquired pneumonia when CRAB was the main pathogen and treatment was not delayed. Because of Abs isolated in this study was sensitive to tigecycline (Mean MIC = 1.5 mg/L) in vitro, and the therapeutic effect of tigecycline might be related to drug sensitivity. Even if these patients had higher APACHE II scores, some of them were prescribed tigecycline alone, there were still good clinical outcomes, if tigecycline was sensitive in vitro.\nMost MDR-AB produce carbapenemases in previous study [13]. We also found that the resistance rate of carbapenem reached 94.6%. Continued use of carbapenem did not reduce the mortality of patients with CRAB pneumonia. It was not recommended to use only carbapenem for theses patients, but it might be effective as one of the combined treatment options, especially the MIC of carbapenem was less than 8. It was worth noting that, the combination of other drugs may reduce the carbapenem MIC value. The most common combination was sulbactam and or tigecycline in clinical practice. Carbapenem and sulbactam had been shown to have a synergistic effect and could reduce mortality [14]. Sulbactam, a β-lactamase inhibitor, has a suitable clinical therapeutic effect, and most of Abs has a low rate of resistance to sulbactam. Sulbactam enhanced the antibacterial effect of carbapenem by acting on the penicillin-binding protein. A previous study suggested that, compared with a single drug, carbapenem combined with sulbactam enhanced the antibacterial effect on CRAB in vitro and decreased the MIC value of carbapenem [15]. Our study found that although Abs was resistant to carbapenem, the 28-day mortality rate of patients after the addition of sulbactam was significantly reduced, which might be related to a mechanism by which sulbactam reduced the MIC value of carbapenem.\nIn addition, the combined dose of sulbactam needs to be further explored. A previous study found that when patients were given sulbactam 2 g q6h or 3 g q8h intravenously, the T > MIC (above 60%) was 98.83% and 95.59% respectively, if the sulbactam MIC was 16 µg/ml. Therefore, high-dose sulbactam was beneficial to controlling MDR-AB infection [16]. Another study recommended that a high-dose sulbactam (9 g/d) was also beneficial in controlling MDR-AB infection [17]. Our study found that the mortality rate of patients with CRAB pneumonia was lower in the high-dose sulbactam group as compared with the low-dose group, but the difference was not statistically significant. This might be related to the small number of cases in the high-dose sulbactam group. Thus, future studies regarding the dose of sulbactam need to include a larger number of cases.\nOur research has many limitations. First, we followed 1,823 clinical cases with positive AB culture results. Most of the cases were excluded because clinicians were not sure whether AB was the causative bacterium of severe pneumonia or only colonization in the lower respiratory tract. Especially at the same time, other bacterial and or fungal pathogens were cultivated from the lower respiratory tract. Second, only 166 cases were identified as severe MDR-AB pneumonia. After grouping, some data results did not reach statistical differences, for example, comparison of the clinical effects of high- and low-dose sulbactam for CRAB. Then, further mechanistic studies were needed, including the paper diffusion method or the antibiotic concentration gradient method (E-test method), to detect whether there was a synergistic effect between carbapenem and sulbactam, and different resistance mechanism in these CRAB strains.\nIn summary, despite the above shortcomings, ICU status, an extended period of invasive catheter retention time and mechanical ventilation time, and not using the appropriate antibacterial treatment increased the mortality in patients with severe hospital-acquired MDR-AB pneumonia. Although MDR-AB was resistant to many antibiotics, most of the MDR-AB strains in western China were still sensitive to tigecycline. The inclusion of tigecycline in the combined treatment plan could significantly reduce the 28-day mortality of patients. In addition, the inclusion of sulbactam in the treatment plan could also reduce the mortality of patients with carbapenemase-producing AB pneumonia on the basis of tigecycline. High-dose sulbactam might improve the clinical prognosis of patients by reducing the MIC of carbapenem drugs especially."],"string":"[\n \"Acinetobacter baumannii (AB) is a common pathogen that causes opportunistic infections in hospitals. According to the China Antimicrobial Surveillance Network (CHINET) continuous drug resistance monitoring program, AB is the second most common gram-negative bacterium isolated from respiratory specimens and accounted for about 17.07% of infections in 2020. Important risk factors for AB infection include admission to the ICU, impaired immune function, receiving broad-spectrum antibacterial drugs, and mechanical ventilation [1,2]. Since AB has a high rate of resistance to antimicrobial drugs, the control of multi-drug resistant AB (MDR-AB) is currently one of the main goals of nosocomial infection prevention and control.\\nIn Asia, the sensitivity of AB to carbapenems were less than 27% [3]. As such, a number of guidelines recommend the combination of polymyxin, tigecycline, carbapenems, sulbactam, or other antibacterial drugs for the treatment of MDR-AB pneumonia in China or abroad [4, 5]. However, the sensitivity of AB to certain drugs differs in different regions. Thus, it is necessary to design an appropriate combination drug regimen according to the local drug resistance situation and common guidelines and recommendations.\\nIt is a challenge for clinicians to choose antibacterial drugs when trying to control severe MDR-AB infection, especially carbapenemase-resistant Acinetobacter baumannii (CRAB). According to an in vitro drug sensitivity test, the resistance rate of AB to carbapenems ranged from 62–100% [3], and the resistance rate to cefoperazone/sulbactam was 46.3% [6]. Antibiotics containing sulbactam were the main drugs for the treatment of Acinetobacter infection in China, partly because of the economic burden. The conventional dose is 4 g/d for MDR-AB, but it can be increased to 6.0 g/d or even 8.0 g/d according to expert consensus on the diagnosis, treatment, prevention, and control of AB infection in China [4]. In addition to sulbactam, tigecycline is the most used for the control of hospital-acquired MDR-AB, but its clinical efficacy is controversial [7]. Studies have shown that even if AB was sensitive to tigecycline in vitro, it could not improve the prognosis of patients with MDR-AB. The main objective of this study was to investigate feasible and effective treatment options for patients with severe MDR-AB pneumonia, from a clinical perspective. Here, we retrospectively identified factors influencing mortality in patients with severe MDR-AB pneumonia and appropriate antibiotic regimens, especially the combined effect of sulbactam and tigecycline.\",\n \"[SUBTITLE] Subjects [SUBSECTION] We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\nWe enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n[SUBTITLE] Method for detecting carbapenemase-producing genotype [SUBSECTION] The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\\nThe carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\\n[SUBTITLE] Clinical information [SUBSECTION] For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\\nFor each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\\n[SUBTITLE] Treatment [SUBSECTION] Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\nEach treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n[SUBTITLE] Statistical methods [SUBSECTION] SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\\nSPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\",\n \"We enrolled 1,823 respiratory tract specimens cultured as MDR-AB from 2016 to 2020 in this study. Out of the 1,823 cases, 166 patients were diagnosed with MDR-AB pneumonia. The screening process was shown in Fig. 1. This retrospective study was approved by the ethics committee of Sichuan Provincial People’s Hospital (Affiliated Hospital of the University of Electronic Science and Technology of China). Approval number: 2019059B.\\n\\nFig. 1MDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nMDR-AB inclusion and exclusion flow chart. *Severe immunodeficiency patients: long-term use of immunosuppressants, solid organ transplant patients, recent use of high-dose methylprednisolone drugs (more than 280 mg/w), hematological malignancies, and patients with severe agranulocytosis. MDR-AB, Multi-drug resistant Acinetobacter baumannii\",\n \"The carbapenemase detection kits (LFA) was used to detect the 157 CRAB strains, which provided by Beijing Gold Mountainriver Tech Development Co., Ltd. Take the strains out of the − 80 °C freezer 24 h in advance and wait at room temperature to thaw. Dip the mixed strain with an inoculating loop and inoculate it on a blood plate or a MacConkey plate in a four-district line. Place these plates in a constant temperature incubator at 35 °C for 18-24 h. Then, fresh strains of AB grown uniformly over 3 zones on the plates were used for the assays.\",\n \"For each patient, the following information was collected: age, sex, underlying diseases, initial diagnosis at admission, clinical features, laboratory findings, other microorganism expect AB, radiological findings, clinical course until diagnosis, antibiotic and invasive treatment, Acute Physiology and Chronic Health Evaluation II (APACHE II) score at admission, and in-hospital outcome.\",\n \"Each treatment plan used to treat patients with MDR-AB was collected. Treatment plans included invasive catheterization (tracheal intubation, central venous catheterization, and urinary tract catheterization) and antibiotic use (polymyxin, tigecycline, carbapenem, sulbactam, aminoglycosides, quinolones, and tetracyclines). Preparations containing single sulbactam were divided into a high-dose group (> 3 g/day) and a low-dose group (≤ 3 g/day). Cases with delayed treatment were excluded. Delayed treatment meant that the clinician obtained the MDR-AB culture results from the laboratory but did not give effective antimicrobial treatment within 3 days.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.Carbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.Sulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\\n\\nDose and duration of tigecycline: the first dose of tigecycline was 100 mg (2 bottles), and then 50 mg (1 bottle) every 12 h. The intravenous infusion time of tigecycline should be once every 12 h, about 30 ~ 60 min each time. The course of treatment in this study was 5–14 days.\\nCarbapenem uses meropenem, dose and duration: it was prepared with sterile water for injection, containing 250 mg meropenem per 5ml, the concentration was about every 50 mg/ml, and it was administered once every 8 h, 500 mg each time, intravenous drip.\\nSulbactam preparation used single sulbactam to participate in the combined treatment, dose and duration: 2 ~ 4 g per day. The same amount was given intravenously concomitant infusion with cefoperazone once every 12 h or 8 h for 30 ~ 60 min.\",\n \"SPSS 23.0 software (SPSS Inc., Chicago, United States) was used to analyze the data. Measurement-type data were tested for normality and homogeneity of variance. The data that conformed to normal distribution and homogeneity of variance were expressed as the mean ± standard deviation (x ± s) and analyzed using a t-test. An adjusted t-test method was used for the data that conformed to normal distribution but did not conform to homogeneity of variance. The data that did not conform to normal distribution were expressed as the median [M (Ql, Qu)], and the rank sum test was used for analysis. A chi-square test was used for counting data. After statistical analysis, the independent variables with statistical significance were selected, and the selected variables were subjected to multi-independent logistic regression analysis. A p-value < 0.05 was regarded as statistically significant.\",\n \"[SUBTITLE] General clinical conditions [SUBSECTION] Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\\n\\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nClinical data of patients with MDR-AB pneumonia\\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\nOf the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\\n\\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nClinical data of patients with MDR-AB pneumonia\\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n[SUBTITLE] Characteristics of Acinetobacter baumannii strains(abs) [SUBSECTION] Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\\n\\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\\n\\nThe susceptibility results of Abs in vitro\\n*Carbapenems including Imipenem/Meropenem/Doripenem\\nOf the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\\n\\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\\n\\nThe susceptibility results of Abs in vitro\\n*Carbapenems including Imipenem/Meropenem/Doripenem\\n[SUBTITLE] Antibiotic medication [SUBSECTION] All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\\n\\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\n\\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\nAll patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\\n\\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\n\\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\n[SUBTITLE] Tigecycline [SUBSECTION] Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\\nAmong the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\\n[SUBTITLE] Carbapenem [SUBSECTION] There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\\nThere were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\\n[SUBTITLE] Sulbactam [SUBSECTION] From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\\n\\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\nFrom the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\\n\\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n[SUBTITLE] Colistin [SUBSECTION] There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\\nThere were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\",\n \"Of the 166 patients diagnosed with severe hospital-acquired MDR-AB pneumonia, 111 were males (Table 1). The median age was 69 years (19–94 years). We divided the patients with severe MDR-AB pneumonia into two groups based on the clinical outcome during 28 days of treatment: (1) death group (n = 45) and (2) survival group (n = 112). The death group was mainly concentrated in the age group over 75 years old (73.3%), which was consistent with our analysis. That patients older than 75 years had an increased risk of death (OR 0.921, p = 0.001). The median APACHE II score of the study population was 20. (The APACHE II score was used to assess the condition and prognosis of ICU patients. A score of > 15 was classified as more critically ill. The higher the score, the higher the risk of death during hospitalization.) Common underlying diseases were neurological diseases, heart diseases, and chronic lung diseases. Mechanical ventilation (OR 4.525) and invasive catheterization (OR 48.526) significantly increased the risk of death in severe MDR-AB patients (p < 0.05).\\n\\nTable 1Clinical data of patients with MDR-AB pneumoniaDeath group(n = 45)Survival group(n = 121)P valueDataAge (Y)81 (72.5–85.5)64 (50–75.5)< 0.001Gender29 (64.4%)82 (67.8%)0.686APACHEII Scores20 (20, 22.5)20 (16, 33)0.094Underlying diseasesLiver diseases7 (15.6%)12 (9.9%)0.31Neurovascular disease18 (40.0%)53 (43.8%)0.66Kidney diseases16 (35.6%)14 (11.6%)< 0.001Diabetes11 (24.4%)16 (13.2%)0.082Pulmonary diseases23 (51.1%)37 (30.6%)0.014Heart disease27 (60.0%)41 (33.9%)0.002Immune abnormalities4 (8.9%)14 (11.6%)0.621TreatmentTreatment delay7 (15.6%)19 (15.7%)0.982Mechanical ventilation38 (84.4%)65 (53.7%)< 0.001Invasive catheterization44 (95.6%)77 (63.6%)< 0.001Carbapenems43 (95.6%)92 (76.0%)0.004Tigecycline20 (44.4%)98 (81.0%)< 0.001Sulbactam20 (44.4%)80 (66.1%)0.011APACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\\n\\nClinical data of patients with MDR-AB pneumonia\\nAPACHEII, Acute Physiology and Chronic Health Evaluation II; MDR-AB, Multi-drug resistant Acinetobacter baumannii\",\n \"Of the 166 patients diagnosed with severe MDR-AB, 157 were confirmed to be carbapenemase-producing. OXA-23 accounted for 98.1% of the genotypes of these CRAB strains, and only 3 patients had NDM. The drug susceptibility results of all Abs in vitro were shown in Table 2.\\n\\nTable 2The susceptibility results of Abs in vitroTotal(n = 166)Death group(n = 45)Survival group(n = 121)SISISIAmpicillin/sulbactam3/1668/1661/450/452/1218/121Piperacillin0/1660/1660/450/450/1210/121Carbapenems*9/1660/1664/450/455/1210/121Piperacillin/Tazobactam1/1663/1660/450/451/1213/121Ceftriaxone0/1664/1660/451/450/1213/121Cefotaxime0/1664/1660/451/450/1213/121Cefepime2/1663/1661/451/451/1212/121Cefoxitin0/1660/1660/450/450/1210/121Gentamicin7/1664/1662/451/455/1213/121Tobramycin15/1662/1663/451/4512/1211/121Amikacin12/1661/1663/450/459/1211/121Ciprofloxacin4/1660/1661/450/453/1210/121Levofloxacin4/1667/1661/450/453/1217/121Moxifloxacin4/1660/1661/450/453/1210/121Tetracycline3/1666/1661/450/452/1216/121Tigecycline161/1665/16644/451/45117/1214/121Trimethoprim/Sulfamethoxazole17/1660/1665/450/4512/1210/121Cefoperazone/Sulbactam3/1667/1661/452/452/1215/121Minocyclline25/16613/1664/452/4521/12111/121Ampicillin1/1660/1660/450/451/1210/121Amoxicillin/Clavulanic6/1660/1662/450/454/1210/121Aztreonam7/1660/1663/450/454/1210/121Cefotetan1/1660/1661/450/450/1210/121Cefazolin0/1660/1660/450/450/1210/121*Carbapenems including Imipenem/Meropenem/Doripenem\\n\\nThe susceptibility results of Abs in vitro\\n*Carbapenems including Imipenem/Meropenem/Doripenem\",\n \"All patients were prescribed 2 or more antibiotics for CRAB pneumonia. According to the recommendations of the MDR-AB diagnosis and treatment guidelines, most of the patients in the study used a combination treatment plan, and only a few were treated with one drug. Some single-agent treatments were explained in detail below (details were shown in Table 3).\\n\\nTable 3Antibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumoniaAntibioticsNumber of casesNumber of deaths(Deaths/Total*100%)Sulbactam + Carbapenems + Tigecycline5912(20.34%)Sulbactam + Carbapenems243(1.25%)Sulbactam + Tigecycline302(6%)Carbapenems + Tigecycline233(13.04%)Sulbactam00(-)Carbapenems*2121(100%)Colistin21(50%)Not using any of the above three drugs00(-)* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\\n\\nAntibiotic usage and clinical outcomes in patients with CRAB pneumonia. Clinical outcomes of different antibiotic regimens in CRAB pneumonia\\n* These patients were treated with other types of antibiotics, including 12 cases with quinolones, 7 cases with other β-lactams, and 2 cases with tetracyclines\",\n \"Among the 166 patients diagnosed with sever MDR-AB pneumonia, the resistance rate to tigecycline was 3% (5/166) in vitro ( Sensitivity to tigecycline was defined as MIC < 2 mg/L). There were 113 cases were prescribed tigecycline, and the 28-day mortality rate was 16.7%. The remaining 48 patients who did not use tigecycline exhibited a higher mortality rate of 53.2% (p < 0.001). There were another five patients were not prescribed tigecycline because of the MIC values of tigecycline higher than 2.\",\n \"There were 157 cases resistant to carbapenem, of which 127 cases produced carbapenemase. These patients were diagnosed with CRAB pneumonia. The resistance rate to carbapenem was 94.6% (157/166). The enzyme-producing genotypes of these CRABs have been described previously. Except for 3 strains producing NDM, the rest were all OXA-23 type. Although the carbapenem resistance rate was high, there were still 104 cases with carbapenem included in the combination treatment regimen. Based on whether tigecycline (T) or sulbactam (S) was used, they were divided into TS group and TS unused group. As a result, the mortality rate in TS group was 6.7%; however, the mortality rate in TS unused group was 30.7% (p = 0.007). The main combined antibiotics in the TS unused group included quinolones (12 cases), other β-lactam antibiotics (9 cases), and tetracyclines (2 cases).\",\n \"From the above results, sulbactam was effective in the treatment of CRAB in addition to tigecycline. The 28-day mortality of patients with CRAB pneumonia was completely different among the different combination regimens, as shown in Fig. 2. The mortality rate of the regimens that included tigecycline, carbapenem, and sulbactam was much lower than that of tigecycline plus carbapenem alone. The effect of the dose of sulbactam on the efficacy of CRAB should be further clarified. The mortality rate in CRAB pneumonia patients treated with carbapenem combined with sulbactam was 18.1%, whereas the mortality rate was 54.5% (p < 0.001) in those patients who were treated with the combination of carbapenem and other antimicrobial drugs (including aminoglycosides, fluoroquinolones, and tetracyclines). For CRAB pneumonia, the addition of sulbactam may reduce the MIC value of carbapenem through a synergistic effect. These patients who were given carbapenem and sulbactam in combination could be divided into a high-dose group and a low-dose group according to the sulbactam dosage. The high-dose group (30 cases) had a mortality rate of 16.7%, and the low-dose group (53 cases) had a mortality rate of 18.9% (p = 0.802) (Fig. 3).\\n\\nFig. 2The 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nThe 28-day survival rate in CRAB patients treated with Carbapenem + Tigecycline or Carbapenem + Tigecycline and Sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nFig. 3Mortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\\n\\nMortality rate in patients with CRAB pneumonia treated with Carbapenem + Tigecycline alone or Carbapenem + Tigecycline and sulbactam. CRAB, carbapenem-resistant Acinetobacter baumannii\",\n \"There were only 2 patients received colistin for the treatment of CRAB. Patient 1 was automatically discharged from the hospital with colistin for less than 3 days; The second patient had delayed treatment and was diagnosed with invasive Aspergillus pulmonary pneumonia at the same time. Therefore, there were no sufficient data to be included in the evaluation of efficacy regarding colistin or combination regimens of colistin.\",\n \"MDR-AB increases the mortality rate of nosocomial pneumonia. Our study found that admission to the intensive care unit, age > 75 years, invasive catheterization, and mechanical ventilation increased the mortality of patients with severe MDR-AB pneumonia, which was consistent with the results of other studies [9]. Early removal of invasive catheters and reducing the time on mechanical ventilation might reduce the mortality of critically ill patients diagnosed with severe MDR-AB pneumonia. Table 1 shows that there were age differences in clinical outcomes in the study population, but no differences by gender. This may be related to factors such as more complications, more co-morbidities, difficulty in weaning from ventilator, and longer ICU stay in older adults with severe pneumonia.\\nIn our study, the main drugs that affect the mortality of patients with severe MDR-AB pneumonia or CRAB pneumonia included tigecycline, carbapenem, and sulbactam. Although Abs had good sensitivity to colistin in vitro, the MIC of colistin was less than 1 in China. But colistin was not widely used in the treatment of CRAB because of its high cost and not being covered by general health insurance in our country. There were only 2 patients were treated with colistin for CRAB, and they could not be included in the study due to uncertainty of clinical data and co-infection of fungal.\\nAlthough tigecycline was recommended as a first-line drug for CRAB pneumonia, there were different opinions on whether its use was beneficial to reduce mortality. The resistance rate of tigecycline in vitro was low in China (about 2.9%) [6], which ranges from 0.2 to 74.2% in other countries and regions [3]. The concentration of tigecycline was higher in tissue, and it has suitable antibacterial activity and was safe for use against local tissue infections. Therefore, it has been theorized that tigecycline was an effective antibiotic for the treatment of CRAB and can reduce mortality [10. A previous meta-analysis showed that patients treated with tigecycline for CRAB infection had a higher mortality rate, longer hospital stay, and lower microbial clearance rate as compared with other drugs [11]. Another prospective study demonstrated that tigecycline increases the mortality rate in patients with CRAB infection when the tigecycline MIC > 2 mg/L [12]. The results of our study showed that tigecycline could significantly reduce the mortality rate of hospital-acquired pneumonia when CRAB was the main pathogen and treatment was not delayed. Because of Abs isolated in this study was sensitive to tigecycline (Mean MIC = 1.5 mg/L) in vitro, and the therapeutic effect of tigecycline might be related to drug sensitivity. Even if these patients had higher APACHE II scores, some of them were prescribed tigecycline alone, there were still good clinical outcomes, if tigecycline was sensitive in vitro.\\nMost MDR-AB produce carbapenemases in previous study [13]. We also found that the resistance rate of carbapenem reached 94.6%. Continued use of carbapenem did not reduce the mortality of patients with CRAB pneumonia. It was not recommended to use only carbapenem for theses patients, but it might be effective as one of the combined treatment options, especially the MIC of carbapenem was less than 8. It was worth noting that, the combination of other drugs may reduce the carbapenem MIC value. The most common combination was sulbactam and or tigecycline in clinical practice. Carbapenem and sulbactam had been shown to have a synergistic effect and could reduce mortality [14]. Sulbactam, a β-lactamase inhibitor, has a suitable clinical therapeutic effect, and most of Abs has a low rate of resistance to sulbactam. Sulbactam enhanced the antibacterial effect of carbapenem by acting on the penicillin-binding protein. A previous study suggested that, compared with a single drug, carbapenem combined with sulbactam enhanced the antibacterial effect on CRAB in vitro and decreased the MIC value of carbapenem [15]. Our study found that although Abs was resistant to carbapenem, the 28-day mortality rate of patients after the addition of sulbactam was significantly reduced, which might be related to a mechanism by which sulbactam reduced the MIC value of carbapenem.\\nIn addition, the combined dose of sulbactam needs to be further explored. A previous study found that when patients were given sulbactam 2 g q6h or 3 g q8h intravenously, the T > MIC (above 60%) was 98.83% and 95.59% respectively, if the sulbactam MIC was 16 µg/ml. Therefore, high-dose sulbactam was beneficial to controlling MDR-AB infection [16]. Another study recommended that a high-dose sulbactam (9 g/d) was also beneficial in controlling MDR-AB infection [17]. Our study found that the mortality rate of patients with CRAB pneumonia was lower in the high-dose sulbactam group as compared with the low-dose group, but the difference was not statistically significant. This might be related to the small number of cases in the high-dose sulbactam group. Thus, future studies regarding the dose of sulbactam need to include a larger number of cases.\\nOur research has many limitations. First, we followed 1,823 clinical cases with positive AB culture results. Most of the cases were excluded because clinicians were not sure whether AB was the causative bacterium of severe pneumonia or only colonization in the lower respiratory tract. Especially at the same time, other bacterial and or fungal pathogens were cultivated from the lower respiratory tract. Second, only 166 cases were identified as severe MDR-AB pneumonia. After grouping, some data results did not reach statistical differences, for example, comparison of the clinical effects of high- and low-dose sulbactam for CRAB. Then, further mechanistic studies were needed, including the paper diffusion method or the antibiotic concentration gradient method (E-test method), to detect whether there was a synergistic effect between carbapenem and sulbactam, and different resistance mechanism in these CRAB strains.\\nIn summary, despite the above shortcomings, ICU status, an extended period of invasive catheter retention time and mechanical ventilation time, and not using the appropriate antibacterial treatment increased the mortality in patients with severe hospital-acquired MDR-AB pneumonia. Although MDR-AB was resistant to many antibiotics, most of the MDR-AB strains in western China were still sensitive to tigecycline. The inclusion of tigecycline in the combined treatment plan could significantly reduce the 28-day mortality of patients. In addition, the inclusion of sulbactam in the treatment plan could also reduce the mortality of patients with carbapenemase-producing AB pneumonia on the basis of tigecycline. High-dose sulbactam might improve the clinical prognosis of patients by reducing the MIC of carbapenem drugs especially.\"\n]"},"all_sections_sec_types":{"kind":"list like","value":[null,null,null,null,null,null,null,"results",null,null,null,null,null,null,null,"discussion"],"string":"[\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"results\",\n null,\n null,\n null,\n null,\n null,\n null,\n null,\n \"discussion\"\n]"},"keywords":{"kind":"list like","value":["Multi-drug resistant Acinetobacter baumannii (MDR-AB)","Carbapenem-resistant Acinetobacter baumannii(CRAB)","Severe pneumonia","Intensive care unit"],"string":"[\n \"Multi-drug resistant Acinetobacter baumannii (MDR-AB)\",\n \"Carbapenem-resistant Acinetobacter baumannii(CRAB)\",\n \"Severe pneumonia\",\n \"Intensive care unit\"\n]"}}},{"rowIdx":391,"cells":{"title":{"kind":"string","value":"Neutrophil activation and NETosis are the predominant drivers of airway inflammation in an OVA/CFA/LPS induced murine model."},"pmid":{"kind":"string","value":"36271366"},"background_abstract":{"kind":"string","value":"Asthma is one of the most common chronic diseases that affects more than 300 million people worldwide. Though most asthma can be well controlled, individuals with severe asthma experience recurrent exacerbations and impose a substantial economic burden on healthcare system. Neutrophil inflammation often occurs in patients with severe asthma who have poor response to glucocorticoids, increasing the difficulty of clinical treatment."},"background_abstract_label":{"kind":"string","value":"BACKGROUND"},"methods_abstract":{"kind":"string","value":"We established several neutrophil-dominated allergic asthma mouse models, and analyzed the airway hyperresponsiveness, airway inflammation and lung pathological changes. Neutrophil extracellular traps (NETs) formation was analyzed using confocal microscopy and western blot."},"methods_abstract_label":{"kind":"string","value":"METHODS"},"results_abstract":{"kind":"string","value":"We found that the ovalbumin (OVA)/complete Freund's adjuvant (CFA)/low-dose lipopolysaccharide (LPS)-induced mouse model best recapitulated the complex alterations in the airways of human severe asthmatic patients. We also observed OVA/CFA/LPS-exposed mice produced large quantities of neutrophil extracellular traps (NETs) in lung tissue and bone marrow neutrophils. Furthermore, we found that reducing the production of NETs or increasing the degradation of NETs can reduce airway inflammation and airway hyperresponsiveness."},"results_abstract_label":{"kind":"string","value":"RESULTS"},"conclusions_abstract":{"kind":"string","value":"Our findings identify a novel mouse model of neutrophilic asthma. We have also identified NETs play a significant role in neutrophilic asthma models and contribute to neutrophilic asthma pathogenesis. NETs may serve as a promising therapeutic target for neutrophilic asthma."},"conclusions_abstract_label":{"kind":"string","value":"CONCLUSION"},"mesh_descriptor_names":{"kind":"list like","value":["Mice","Humans","Animals","Ovalbumin","Lipopolysaccharides","Neutrophil Activation","Freund's Adjuvant","Disease Models, Animal","Glucocorticoids","Asthma","Inflammation","Respiratory Hypersensitivity"],"string":"[\n \"Mice\",\n \"Humans\",\n \"Animals\",\n \"Ovalbumin\",\n \"Lipopolysaccharides\",\n \"Neutrophil Activation\",\n \"Freund's Adjuvant\",\n \"Disease Models, Animal\",\n \"Glucocorticoids\",\n \"Asthma\",\n \"Inflammation\",\n \"Respiratory Hypersensitivity\"\n]"},"pmcid":{"kind":"string","value":"9587569"},"background_title":{"kind":"string","value":"Introduction"},"background_text":{"kind":"string","value":"